34
Environment and Development Economics 15: 241–274 C Cambridge University Press 2010 doi:10.1017/S1355770X1000001X Finance for renewable energy: an empirical analysis of developing and transition economies CHRISTA N. BRUNNSCHWEILER CER-ETH – Center of Economic Research at ETH Zurich, Zuerichbergstrasse 18, 8092 Zurich, Switzerland and OxCarre, University of Oxford. Email: [email protected] Submitted February 4, 2009; revised August 12, 2009; accepted January 2, 2010; first published online 26 February 2010 ABSTRACT. This paper examines the role of the financial sector in renewable energy (RE) development. Although RE can bring socio-economic and environmental benefits, its implementation faces a number of obstacles, especially in non-OECD countries. One of these obstacles is financing: underdeveloped financial sectors are unable to efficiently channel loans to RE producers. The influence of financial sector development on the use of renewable energy resources is confirmed in panel data estimations on up to 119 non-OECD countries for 1980–2006. Financial intermediation, in particular commercial banking, has a significant positive effect on the amount of RE produced, and the impact is especially large when we consider non-hydropower RE such as wind, solar, geothermal and biomass. There is also evidence that the development of the RE sector has picked up significantly in the period since the adoption of the Kyoto Protocol. 1. Introduction Achieving a diversified and sustainable energy supply for future generations is one of the major challenges for today’s policymakers. Global energy demand is projected to grow by around 45 per cent by 2030: more than three-quarters of the increased demand will come from developing and transition countries (IEA, 2008). Energy demand will continue to be covered mainly by conventional fossil fuels, such as coal, oil and natural gas; accordingly, energy-related pollution is predicted to increase by up to 45 per cent. Although Organization for Economic Co-operation and Development (OECD) countries will still be major polluters, 97 per cent of the estimated increase will come from non-OECD countries, especially China, India and the Middle East (IEA, 2008). Meanwhile, many estimates predict that oil This is a fully revised version of a previous working paper titled ‘Financing the alternative: renewable energy in developing and transition countries’ (Brunnschweiler, 2006). I am grateful to the members of the Resource Economics group at CER-ETH, to participants in the 2009 EAERE conference, and to an Associate Editor and three anonymous referees for useful suggestions. I also thank Luca Deidda, Karen Pittel, Simone Valente, Axel Wieneke, the members of the Center for Corporate Responsibility and Sustainability (CCRS) of the University of Zurich for their helpful comments on the previous version. terms of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S1355770X1000001X Downloaded from https:/www.cambridge.org/core. University of Basel Library, on 11 Jul 2017 at 15:59:45, subject to the Cambridge Core

Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 15 241ndash274 Ccopy Cambridge University Press 2010doi101017S1355770X1000001X

Finance for renewable energy an empiricalanalysis of developing and transition economies

CHRISTA N BRUNNSCHWEILERCER-ETH ndash Center of Economic Research at ETH ZurichZuerichbergstrasse 18 8092 Zurich Switzerland and OxCarre University ofOxford Email cbrunnschweilerethzch

Submitted February 4 2009 revised August 12 2009 accepted January 2 2010first published online 26 February 2010

ABSTRACT This paper examines the role of the financial sector in renewable energy(RE) development Although RE can bring socio-economic and environmental benefitsits implementation faces a number of obstacles especially in non-OECD countries Oneof these obstacles is financing underdeveloped financial sectors are unable to efficientlychannel loans to RE producers The influence of financial sector development on theuse of renewable energy resources is confirmed in panel data estimations on up to 119non-OECD countries for 1980ndash2006 Financial intermediation in particular commercialbanking has a significant positive effect on the amount of RE produced and the impact isespecially large when we consider non-hydropower RE such as wind solar geothermaland biomass There is also evidence that the development of the RE sector has picked upsignificantly in the period since the adoption of the Kyoto Protocol

1 IntroductionAchieving a diversified and sustainable energy supply for futuregenerations is one of the major challenges for todayrsquos policymakers Globalenergy demand is projected to grow by around 45 per cent by 2030 morethan three-quarters of the increased demand will come from developingand transition countries (IEA 2008) Energy demand will continue to becovered mainly by conventional fossil fuels such as coal oil and natural gasaccordingly energy-related pollution is predicted to increase by up to 45 percent Although Organization for Economic Co-operation and Development(OECD) countries will still be major polluters 97 per cent of the estimatedincrease will come from non-OECD countries especially China India andthe Middle East (IEA 2008) Meanwhile many estimates predict that oil

This is a fully revised version of a previous working paper titled lsquoFinancingthe alternative renewable energy in developing and transition countriesrsquo(Brunnschweiler 2006) I am grateful to the members of the Resource Economicsgroup at CER-ETH to participants in the 2009 EAERE conference and to anAssociate Editor and three anonymous referees for useful suggestions I also thankLuca Deidda Karen Pittel Simone Valente Axel Wieneke the members of theCenter for Corporate Responsibility and Sustainability (CCRS) of the Universityof Zurich for their helpful comments on the previous version

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

242 Christa N Brunnschweiler

and possibly natural gas production will plateau around the same timecasting doubt on future energy security1 Hence achieving a sustainableenergy supply requires diversifying energy sources and changing thecurrent dependence on non-renewable and polluting hydrocarbon fuelsFor example in a recent report the UN Commission on SustainableDevelopment states

lsquoEnergy is crucial for sustainable development poverty eradication andachieving the internationally agreed development goals including theMillennium Development Goals [ ] Access to reliable affordableeconomically viable socially acceptable and environmentally sound energyservices is crucial particularly in developing countries [ ] While fossil fuelswill continue to play an important role in the energy supply in the decades tocome every effort must be made to diversify the energy mixrsquo (UN 2007 15)

Renewable energy technologies (RETs) can bring about bothenvironmental and socio-economic benefits2 They generally entail feweremissions use local resources ndash including labour ndash foster basic electrificationin developing countries and increase energy security3 However althoughthere are already several commercially available and economicallyattractive RETs they still account for only a modest proportion of globalenergy generation This fact suggests that there are some missing linksbetween the potential of RETs and their implementation One problemregards the institutional framework and the absence of a policy designto effectively foster RETs In this paper we focus on another importantmissing link the financing of renewable energy (RE) projects in particularthe relationship between financial sector and RET development in transitionand developing countriesThis missing link has been pointed out bynumerous practitioners who see the absence of well-developed financialintermediaries and the consequent financing difficulties as one of the mostimportant obstacles during the realization of RE projects in developingcountries (eg Painuly and Wohlgemuth 2006)

1 The US Energy Information Administration (EIA 2000) alone has publishedseveral different scenarios with global oil production peaking between 2021and 2112 The International Energy Agency (IEA) now expects conventional oilproduction to plateau before 2030 (Economist 2009) Note however that there isalso some skepticism regarding the peak oil scenarios For a discussion of peakoil see for example Deffeyes (2005)

2 RETs include both the more traditional hydropower technologies as well asnewer technologies that harness wind solar biomass and geothermal powerMost recently the use of biofuels and their negative impact on food productionfor example in Brazil has called into question the wisdom of promoting (all typesof) RETs In our investigation we concentrate on electricity generation wherewood and waste are the only types of biomass considered and biofuels thereforeplay no role

3 Barbier (2009) presents a more extensive discussion of the short- to long-termenvironmental social and economic benefits of RETs not only for developed butalso for developing countries One of the effects mentioned is a positive localemployment effect for example in China the RE sector already offers nearly1 million jobs while in India the wind energy sector alone employs around10 thousand people (see also Renner et al 2008)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 243

Energy projects generally demand high levels of financing whichproducers in less developed economies in particular can rarely cover ontheir own (World Bank 1999 IEA 2003) In turn the financing for RETsis closely connected to the development of the financial sector On theone hand energy sector privatisation and liberalisation during the courseof the 1990s increased the contribution of smaller private power projectsand at the same time induced a shift in external financing from the localgovernment and multilateral institutions to private investors (Babbar andSchuster 1998 Tharakan et al 2007 Tirpak and Adams 2008) On the otherhand RE projects have very high start-up costs relative to the expectedmonetary returns and lengthy payback periods they therefore typicallyrequire long-term maturity loans (UNEP FI 2004 Sonntag-OrsquoBrien andUsher 2004b) The result was a plunge in energy project investment inthe mid 1990s as large bilateral and multilateral donors pulled out andinvestment in this sector has struggled to take off again

The problem of financing RE projects is twofold firstly RET firmsgenerally need long-term loans whose availability in turn is positivelylinked to the development of the banking system (Demirguc-Kunt andMaksimovic 1999) In less developed economies the banking sector isthe major source of external financing (Tadesse 2002 Carlin and Mayer2003 Beck et al 2004b) and access to bank credit is a serious problemespecially for small- and medium-sized companies (Beck et al 2004a) Asa consequence RE projects in less developed countries are at a particulardisadvantage Secondly RET firms have limited access to financing becauseRE projects compete against fossil fuel projects which have a longertrack record relatively lower up-front costs shorter lead times and oftenfavourable political treatment (Churchill and Saunders 1989 Head 2000World Bank 2002 Sonntag-OrsquoBrien and Usher 2004b)

It is worth noting that in both cases underinvestment in RET firms can beinterpreted in terms of imperfect information between firms and financiersprojects aimed at developing new technologies bear almost by definitiongreater information costs to investors which are more easily borne by ahighly developed financial sector Where the latter is not given the resultmay well be a market distortion in favour of less risky investments suchas fossil fuel projects and large-sized enterprises This is consistent withthe view that the development of the domestic financial sector is a crucialfactor in meeting the booming energy demand in less developed economies(Ishiguro and Akiyama 1995 World Bank 2003)

RE adoption is one of the targets of the Kyoto Protocol There aretwo mechanisms in particular which can potentially help to overcomethe financing hurdle namely the Clean Development Mechanism (CDM)and the Joint Implementation (JI) programme Both are designed to helpKyoto Protocol member countries ndash in particular Annex B countries ndashmeet their emission targets and to encourage the private sector tocontribute to emission reduction efforts (see eg Pacudan 2005) The CDMmechanism in particular has been quite popular since its inception in2006 with more than 1000 projects already approved It is the only KyotoProtocol mechanism that includes developing countries it allows AnnexB countries that are subject to emission-reduction targets to implement

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

244 Christa N Brunnschweiler

emission-reduction projects in non-Annex B developing countries andearn certified emission reduction credits (JI programmes on the other handinvolve joint projects by two Annex B countries) Some RE projects werealready initiated in the late 1990s as CDM or JI pilot programmes suchas the Tejona wind farm in Costa Rica4 The Kyoto Protocolrsquos adoption inDecember 1997 obviously marks a major change in global climate policy weincorporate the potential lsquoKyoto effectsrsquo into our study to further isolate theimpact of financial sector development on the RE sector However it is stilltoo early to gauge the full impact of the new Kyoto financing possibilitieson the adoption of RE in developing and transition countries this questionis therefore left to future research5

The analysis of the role of the financial sector ndash commercial bankingfinancial markets insurance etc ndash for economic performance has generateda vast literature during the past two decades (see eg Demirguc-Kunt andLevine 2001 for an overview) The importance of the (private) financialsector particularly for the development of the energy sector has also beenpointed out in several studies In one of the earliest analyses of energysector financing in developing countries Churchill and Saunders (1989)discuss a proper policy framework to encourage private sector financialinvolvement Ten years later Babbar and Schuster (1998) and Head (2000)still find substantial gaps in the financing of power projects particularly REprojects The financing obstacles for RETs are confirmed in the overview byWohlgemuth and Painuly (1999) where efforts in different countries andregions are discussed and several policy recommendations derived Morerecently Sonntag-OrsquoBrien and Usher (2004a) and Painuly and Wohlgemuth(2006) take a look at the experience to date with RET implementation indeveloping and transition countries and again point to the (private sector)financing problems that RE projects encounter as well as to some successfulmodels that have been adopted MacLean and Siegel (2007) concentrate onthe financing of small-scale RE projects and distinguish three financingareas end-user finance business finance and small-scale project financeThe need for well-informed (local) commercial financiers often to act as

4 The Tejona wind farm also illustrates the importance of an adequate financingframework The projectrsquos history dates back to the 1970s when the Costa Ricangovernment applied to the regional Development Bank and the World Bank forfinancing assistance A feasibility study followed in 1993 and a private windenergy company sought to develop the wind farm as an independent powerproject However it took until 1998 to find a project financing model which endedup being a build-operate-lease construction together with a Dutch consortiumwith the support of the Dutch government and the coordination of the GlobalEnvironment Facility In the meantime the Costa Rican Electricity Institute haddecided to develop the site as a public sector project The wind farm is nowoperational and has a capacity of 20 MW Several private wind farm developersare negotiating project finance arrangements for new wind power developmentsunder power purchase agreements with the Electricity Institute (see van Hulleet al 2003)

5 See for example Huang and Barker (2009) for a study on the effects to date of theCDM mechanism on CO2 emissions

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 245

financing partners for the government andor international developmentorganizations is a recurrent theme in all three areas6

Despite the importance of the financial sector for the development of RETsparticularly in developing and transition countries which is borne out bynumerous case studies and anecdotal evidence the subject has receivedlittle academic research attention This paper contributes to the knowledgeon what determines RET implementation by empirically analysing therelationship between financial intermediation and RE sector developmentwith a focus on non-OECD countries To the best of our knowledge this isthe first attempt to examine this issue in a systematic empirical analysis7

The literature discussed above suggests that a more highly developedfinancial sector will have a positive impact on the development of theRE sector A well-functioning and unrestricted banking sector should beparticularly relevant for credit allocation to RE companies in developingand transition countries where financial markets are still relatively smalland bond or equity financing is therefore difficult or impossible Of coursea well-developed financial sector alone is no guarantee for the success ofRETs the availability of adequate financing mechanisms should be viewedin the wider context of an appropriate RE policy framework8

We propose an empirical framework to test the effect of financial sectordevelopment on RETs We construct a panel dataset for up to 119 non-OECDcountries for the period 1980ndash2006 using electricity generation per capita

6 Among the examples cited in the report is that of a small hydro project on the WestNile in Uganda The project was realized thanks to a loan from a commercial lenderBarclays which has a strong presence in Africa lsquoGiven their lack of experiencewith this type of project and other financial market conditionsrsquo Barclays washowever only willing to extend a seven-year loan (MacLean and Siegel 200771) In order to make the project affordable the World Bank provided a partialguarantee which extended the loan term to 14 years Note that since that initialexperience Barclays has financed similar projects in Africa including anothersmall hydro plant in Kenya and ndash together with a South African bank ndash a 250 MWhydro-electric station again on the Nile river in Uganda

7 The theoretical contributions on this topic are equally scarce one attempt tomodel the connections between finance and RE sector development is presentedin Brunnschweiler (2006)

8 Missing finance is obviously connected to the more general policy frameworkfor RETs as previous literature has pointed out limited financing of RETsderives not only from underdeveloped financial intermediators but also fromthe lack of a specific policy design andor crowding-out effects from governmentpolicies favouring investment in fossil fuel projects (Churchill and Saunders 1989Wohlgemuth and Painuly 1999 Head 2000 World Bank 2002 Sonntag-OrsquoBrienand Usher 2004b UNEP FI 2004) Institutional shortcomings also contribute to theoften limited consideration by potential investors of the positive environmentalexternalities of RETs in project development costs In general the perceptionthat energy sustainability is not a top priority for policymakers further lowersinvestorsrsquo willingness to finance projects where the foreseeable rewards are alreadyrelatively low and long in the coming (see Williams and Ghanadan 2006 for auseful survey of electricity reform policies in developing and transition countries)We take the policy framework into account in our estimations to isolate the specificeffect of financial sector development (see section 2)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

246 Christa N Brunnschweiler

from RE technologies as a proxy for RE sector development We isolate thefinancial sector effects by controlling for energy-relevant policy measuresas well as measures of more general institutional quality The empiricalresults from generalised least squares (GLS) and dynamic ArellanondashBondGeneralised Method-of-Moments (GMM) panel estimations confirm thepositive effect of financial sector and especially commercial banking sectordevelopment on RET use in developing and transition countries The effectsare particularly strong and economically significant for the newer RETsincluding wind solar geothermal and biomass while the overall impacton RE and hydropower generation is much more limited in magnitudethough still statistically significant

The results also suggest that the adoption of the Kyoto Protocol has hada strong positive effect on the diffusion of RETs throughout developingand transition countries The findings are robust to the inclusion of othercovariates which could influence RE sector development such as oilcoal and natural gas production and prices and proxies for RE potentialThe implication is that without proper (private sector) finance RE isunlikely to reach its full potential in the developing world An adequatefinancing framework should therefore be part of a more general RE sectordevelopment policy

The paper is organised as follows The data and empirical methodologyare described in section 2 while the results are presented in section 3 Section4 concludes with a brief summary and discussion

2 Data and methodologyThe discussion of the role of commercial finance in the development ofRE has so far been based mainly on case studies and anecdotal evidenceThe lack of a more systematic empirical analysis of the correlation betweenfinancial sector and RE development independent of (or complementaryto) a specific RE policy framework has probably also been due to the dataproblem regarding the quantification of the RE sector especially in thedeveloping world

The obstacles begin with the definition of RE in official statisticstraditionally hydropower ndash mostly provided by large plants ndash has deliveredthe lionrsquos share of RE in countriesrsquo energy generation mix with othertypes of RE ndash when included ndash making up barely a few per cent ofoverall energy production Recently however some environmentalists andpolicymakers have contended that large hydropower projects should notbe viewed as viable contributions to sustainable energy production asthey often cause serious negative environmental and social externalities(notable examples are the giant Three Gorges Dam on the Yangtze Riverin China or the Ilisu dam project in Turkey) Moreover most traditionallarge hydro projects in the developing world have been co-financed bymultilateral financial institutions and the local governments with little orno involvement sought of commercial finance (World Bank 2003) Theuse of an overall RE measure could therefore introduce a downwardbias into the results on the importance of the financial sector for moremodern RETs such as wind geothermal and solar power A further possibleissue concerns the negative impact on agricultural (food) production of

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 2: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

242 Christa N Brunnschweiler

and possibly natural gas production will plateau around the same timecasting doubt on future energy security1 Hence achieving a sustainableenergy supply requires diversifying energy sources and changing thecurrent dependence on non-renewable and polluting hydrocarbon fuelsFor example in a recent report the UN Commission on SustainableDevelopment states

lsquoEnergy is crucial for sustainable development poverty eradication andachieving the internationally agreed development goals including theMillennium Development Goals [ ] Access to reliable affordableeconomically viable socially acceptable and environmentally sound energyservices is crucial particularly in developing countries [ ] While fossil fuelswill continue to play an important role in the energy supply in the decades tocome every effort must be made to diversify the energy mixrsquo (UN 2007 15)

Renewable energy technologies (RETs) can bring about bothenvironmental and socio-economic benefits2 They generally entail feweremissions use local resources ndash including labour ndash foster basic electrificationin developing countries and increase energy security3 However althoughthere are already several commercially available and economicallyattractive RETs they still account for only a modest proportion of globalenergy generation This fact suggests that there are some missing linksbetween the potential of RETs and their implementation One problemregards the institutional framework and the absence of a policy designto effectively foster RETs In this paper we focus on another importantmissing link the financing of renewable energy (RE) projects in particularthe relationship between financial sector and RET development in transitionand developing countriesThis missing link has been pointed out bynumerous practitioners who see the absence of well-developed financialintermediaries and the consequent financing difficulties as one of the mostimportant obstacles during the realization of RE projects in developingcountries (eg Painuly and Wohlgemuth 2006)

1 The US Energy Information Administration (EIA 2000) alone has publishedseveral different scenarios with global oil production peaking between 2021and 2112 The International Energy Agency (IEA) now expects conventional oilproduction to plateau before 2030 (Economist 2009) Note however that there isalso some skepticism regarding the peak oil scenarios For a discussion of peakoil see for example Deffeyes (2005)

2 RETs include both the more traditional hydropower technologies as well asnewer technologies that harness wind solar biomass and geothermal powerMost recently the use of biofuels and their negative impact on food productionfor example in Brazil has called into question the wisdom of promoting (all typesof) RETs In our investigation we concentrate on electricity generation wherewood and waste are the only types of biomass considered and biofuels thereforeplay no role

3 Barbier (2009) presents a more extensive discussion of the short- to long-termenvironmental social and economic benefits of RETs not only for developed butalso for developing countries One of the effects mentioned is a positive localemployment effect for example in China the RE sector already offers nearly1 million jobs while in India the wind energy sector alone employs around10 thousand people (see also Renner et al 2008)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 243

Energy projects generally demand high levels of financing whichproducers in less developed economies in particular can rarely cover ontheir own (World Bank 1999 IEA 2003) In turn the financing for RETsis closely connected to the development of the financial sector On theone hand energy sector privatisation and liberalisation during the courseof the 1990s increased the contribution of smaller private power projectsand at the same time induced a shift in external financing from the localgovernment and multilateral institutions to private investors (Babbar andSchuster 1998 Tharakan et al 2007 Tirpak and Adams 2008) On the otherhand RE projects have very high start-up costs relative to the expectedmonetary returns and lengthy payback periods they therefore typicallyrequire long-term maturity loans (UNEP FI 2004 Sonntag-OrsquoBrien andUsher 2004b) The result was a plunge in energy project investment inthe mid 1990s as large bilateral and multilateral donors pulled out andinvestment in this sector has struggled to take off again

The problem of financing RE projects is twofold firstly RET firmsgenerally need long-term loans whose availability in turn is positivelylinked to the development of the banking system (Demirguc-Kunt andMaksimovic 1999) In less developed economies the banking sector isthe major source of external financing (Tadesse 2002 Carlin and Mayer2003 Beck et al 2004b) and access to bank credit is a serious problemespecially for small- and medium-sized companies (Beck et al 2004a) Asa consequence RE projects in less developed countries are at a particulardisadvantage Secondly RET firms have limited access to financing becauseRE projects compete against fossil fuel projects which have a longertrack record relatively lower up-front costs shorter lead times and oftenfavourable political treatment (Churchill and Saunders 1989 Head 2000World Bank 2002 Sonntag-OrsquoBrien and Usher 2004b)

It is worth noting that in both cases underinvestment in RET firms can beinterpreted in terms of imperfect information between firms and financiersprojects aimed at developing new technologies bear almost by definitiongreater information costs to investors which are more easily borne by ahighly developed financial sector Where the latter is not given the resultmay well be a market distortion in favour of less risky investments suchas fossil fuel projects and large-sized enterprises This is consistent withthe view that the development of the domestic financial sector is a crucialfactor in meeting the booming energy demand in less developed economies(Ishiguro and Akiyama 1995 World Bank 2003)

RE adoption is one of the targets of the Kyoto Protocol There aretwo mechanisms in particular which can potentially help to overcomethe financing hurdle namely the Clean Development Mechanism (CDM)and the Joint Implementation (JI) programme Both are designed to helpKyoto Protocol member countries ndash in particular Annex B countries ndashmeet their emission targets and to encourage the private sector tocontribute to emission reduction efforts (see eg Pacudan 2005) The CDMmechanism in particular has been quite popular since its inception in2006 with more than 1000 projects already approved It is the only KyotoProtocol mechanism that includes developing countries it allows AnnexB countries that are subject to emission-reduction targets to implement

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

244 Christa N Brunnschweiler

emission-reduction projects in non-Annex B developing countries andearn certified emission reduction credits (JI programmes on the other handinvolve joint projects by two Annex B countries) Some RE projects werealready initiated in the late 1990s as CDM or JI pilot programmes suchas the Tejona wind farm in Costa Rica4 The Kyoto Protocolrsquos adoption inDecember 1997 obviously marks a major change in global climate policy weincorporate the potential lsquoKyoto effectsrsquo into our study to further isolate theimpact of financial sector development on the RE sector However it is stilltoo early to gauge the full impact of the new Kyoto financing possibilitieson the adoption of RE in developing and transition countries this questionis therefore left to future research5

The analysis of the role of the financial sector ndash commercial bankingfinancial markets insurance etc ndash for economic performance has generateda vast literature during the past two decades (see eg Demirguc-Kunt andLevine 2001 for an overview) The importance of the (private) financialsector particularly for the development of the energy sector has also beenpointed out in several studies In one of the earliest analyses of energysector financing in developing countries Churchill and Saunders (1989)discuss a proper policy framework to encourage private sector financialinvolvement Ten years later Babbar and Schuster (1998) and Head (2000)still find substantial gaps in the financing of power projects particularly REprojects The financing obstacles for RETs are confirmed in the overview byWohlgemuth and Painuly (1999) where efforts in different countries andregions are discussed and several policy recommendations derived Morerecently Sonntag-OrsquoBrien and Usher (2004a) and Painuly and Wohlgemuth(2006) take a look at the experience to date with RET implementation indeveloping and transition countries and again point to the (private sector)financing problems that RE projects encounter as well as to some successfulmodels that have been adopted MacLean and Siegel (2007) concentrate onthe financing of small-scale RE projects and distinguish three financingareas end-user finance business finance and small-scale project financeThe need for well-informed (local) commercial financiers often to act as

4 The Tejona wind farm also illustrates the importance of an adequate financingframework The projectrsquos history dates back to the 1970s when the Costa Ricangovernment applied to the regional Development Bank and the World Bank forfinancing assistance A feasibility study followed in 1993 and a private windenergy company sought to develop the wind farm as an independent powerproject However it took until 1998 to find a project financing model which endedup being a build-operate-lease construction together with a Dutch consortiumwith the support of the Dutch government and the coordination of the GlobalEnvironment Facility In the meantime the Costa Rican Electricity Institute haddecided to develop the site as a public sector project The wind farm is nowoperational and has a capacity of 20 MW Several private wind farm developersare negotiating project finance arrangements for new wind power developmentsunder power purchase agreements with the Electricity Institute (see van Hulleet al 2003)

5 See for example Huang and Barker (2009) for a study on the effects to date of theCDM mechanism on CO2 emissions

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 245

financing partners for the government andor international developmentorganizations is a recurrent theme in all three areas6

Despite the importance of the financial sector for the development of RETsparticularly in developing and transition countries which is borne out bynumerous case studies and anecdotal evidence the subject has receivedlittle academic research attention This paper contributes to the knowledgeon what determines RET implementation by empirically analysing therelationship between financial intermediation and RE sector developmentwith a focus on non-OECD countries To the best of our knowledge this isthe first attempt to examine this issue in a systematic empirical analysis7

The literature discussed above suggests that a more highly developedfinancial sector will have a positive impact on the development of theRE sector A well-functioning and unrestricted banking sector should beparticularly relevant for credit allocation to RE companies in developingand transition countries where financial markets are still relatively smalland bond or equity financing is therefore difficult or impossible Of coursea well-developed financial sector alone is no guarantee for the success ofRETs the availability of adequate financing mechanisms should be viewedin the wider context of an appropriate RE policy framework8

We propose an empirical framework to test the effect of financial sectordevelopment on RETs We construct a panel dataset for up to 119 non-OECDcountries for the period 1980ndash2006 using electricity generation per capita

6 Among the examples cited in the report is that of a small hydro project on the WestNile in Uganda The project was realized thanks to a loan from a commercial lenderBarclays which has a strong presence in Africa lsquoGiven their lack of experiencewith this type of project and other financial market conditionsrsquo Barclays washowever only willing to extend a seven-year loan (MacLean and Siegel 200771) In order to make the project affordable the World Bank provided a partialguarantee which extended the loan term to 14 years Note that since that initialexperience Barclays has financed similar projects in Africa including anothersmall hydro plant in Kenya and ndash together with a South African bank ndash a 250 MWhydro-electric station again on the Nile river in Uganda

7 The theoretical contributions on this topic are equally scarce one attempt tomodel the connections between finance and RE sector development is presentedin Brunnschweiler (2006)

8 Missing finance is obviously connected to the more general policy frameworkfor RETs as previous literature has pointed out limited financing of RETsderives not only from underdeveloped financial intermediators but also fromthe lack of a specific policy design andor crowding-out effects from governmentpolicies favouring investment in fossil fuel projects (Churchill and Saunders 1989Wohlgemuth and Painuly 1999 Head 2000 World Bank 2002 Sonntag-OrsquoBrienand Usher 2004b UNEP FI 2004) Institutional shortcomings also contribute to theoften limited consideration by potential investors of the positive environmentalexternalities of RETs in project development costs In general the perceptionthat energy sustainability is not a top priority for policymakers further lowersinvestorsrsquo willingness to finance projects where the foreseeable rewards are alreadyrelatively low and long in the coming (see Williams and Ghanadan 2006 for auseful survey of electricity reform policies in developing and transition countries)We take the policy framework into account in our estimations to isolate the specificeffect of financial sector development (see section 2)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

246 Christa N Brunnschweiler

from RE technologies as a proxy for RE sector development We isolate thefinancial sector effects by controlling for energy-relevant policy measuresas well as measures of more general institutional quality The empiricalresults from generalised least squares (GLS) and dynamic ArellanondashBondGeneralised Method-of-Moments (GMM) panel estimations confirm thepositive effect of financial sector and especially commercial banking sectordevelopment on RET use in developing and transition countries The effectsare particularly strong and economically significant for the newer RETsincluding wind solar geothermal and biomass while the overall impacton RE and hydropower generation is much more limited in magnitudethough still statistically significant

The results also suggest that the adoption of the Kyoto Protocol has hada strong positive effect on the diffusion of RETs throughout developingand transition countries The findings are robust to the inclusion of othercovariates which could influence RE sector development such as oilcoal and natural gas production and prices and proxies for RE potentialThe implication is that without proper (private sector) finance RE isunlikely to reach its full potential in the developing world An adequatefinancing framework should therefore be part of a more general RE sectordevelopment policy

The paper is organised as follows The data and empirical methodologyare described in section 2 while the results are presented in section 3 Section4 concludes with a brief summary and discussion

2 Data and methodologyThe discussion of the role of commercial finance in the development ofRE has so far been based mainly on case studies and anecdotal evidenceThe lack of a more systematic empirical analysis of the correlation betweenfinancial sector and RE development independent of (or complementaryto) a specific RE policy framework has probably also been due to the dataproblem regarding the quantification of the RE sector especially in thedeveloping world

The obstacles begin with the definition of RE in official statisticstraditionally hydropower ndash mostly provided by large plants ndash has deliveredthe lionrsquos share of RE in countriesrsquo energy generation mix with othertypes of RE ndash when included ndash making up barely a few per cent ofoverall energy production Recently however some environmentalists andpolicymakers have contended that large hydropower projects should notbe viewed as viable contributions to sustainable energy production asthey often cause serious negative environmental and social externalities(notable examples are the giant Three Gorges Dam on the Yangtze Riverin China or the Ilisu dam project in Turkey) Moreover most traditionallarge hydro projects in the developing world have been co-financed bymultilateral financial institutions and the local governments with little orno involvement sought of commercial finance (World Bank 2003) Theuse of an overall RE measure could therefore introduce a downwardbias into the results on the importance of the financial sector for moremodern RETs such as wind geothermal and solar power A further possibleissue concerns the negative impact on agricultural (food) production of

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 3: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 243

Energy projects generally demand high levels of financing whichproducers in less developed economies in particular can rarely cover ontheir own (World Bank 1999 IEA 2003) In turn the financing for RETsis closely connected to the development of the financial sector On theone hand energy sector privatisation and liberalisation during the courseof the 1990s increased the contribution of smaller private power projectsand at the same time induced a shift in external financing from the localgovernment and multilateral institutions to private investors (Babbar andSchuster 1998 Tharakan et al 2007 Tirpak and Adams 2008) On the otherhand RE projects have very high start-up costs relative to the expectedmonetary returns and lengthy payback periods they therefore typicallyrequire long-term maturity loans (UNEP FI 2004 Sonntag-OrsquoBrien andUsher 2004b) The result was a plunge in energy project investment inthe mid 1990s as large bilateral and multilateral donors pulled out andinvestment in this sector has struggled to take off again

The problem of financing RE projects is twofold firstly RET firmsgenerally need long-term loans whose availability in turn is positivelylinked to the development of the banking system (Demirguc-Kunt andMaksimovic 1999) In less developed economies the banking sector isthe major source of external financing (Tadesse 2002 Carlin and Mayer2003 Beck et al 2004b) and access to bank credit is a serious problemespecially for small- and medium-sized companies (Beck et al 2004a) Asa consequence RE projects in less developed countries are at a particulardisadvantage Secondly RET firms have limited access to financing becauseRE projects compete against fossil fuel projects which have a longertrack record relatively lower up-front costs shorter lead times and oftenfavourable political treatment (Churchill and Saunders 1989 Head 2000World Bank 2002 Sonntag-OrsquoBrien and Usher 2004b)

It is worth noting that in both cases underinvestment in RET firms can beinterpreted in terms of imperfect information between firms and financiersprojects aimed at developing new technologies bear almost by definitiongreater information costs to investors which are more easily borne by ahighly developed financial sector Where the latter is not given the resultmay well be a market distortion in favour of less risky investments suchas fossil fuel projects and large-sized enterprises This is consistent withthe view that the development of the domestic financial sector is a crucialfactor in meeting the booming energy demand in less developed economies(Ishiguro and Akiyama 1995 World Bank 2003)

RE adoption is one of the targets of the Kyoto Protocol There aretwo mechanisms in particular which can potentially help to overcomethe financing hurdle namely the Clean Development Mechanism (CDM)and the Joint Implementation (JI) programme Both are designed to helpKyoto Protocol member countries ndash in particular Annex B countries ndashmeet their emission targets and to encourage the private sector tocontribute to emission reduction efforts (see eg Pacudan 2005) The CDMmechanism in particular has been quite popular since its inception in2006 with more than 1000 projects already approved It is the only KyotoProtocol mechanism that includes developing countries it allows AnnexB countries that are subject to emission-reduction targets to implement

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

244 Christa N Brunnschweiler

emission-reduction projects in non-Annex B developing countries andearn certified emission reduction credits (JI programmes on the other handinvolve joint projects by two Annex B countries) Some RE projects werealready initiated in the late 1990s as CDM or JI pilot programmes suchas the Tejona wind farm in Costa Rica4 The Kyoto Protocolrsquos adoption inDecember 1997 obviously marks a major change in global climate policy weincorporate the potential lsquoKyoto effectsrsquo into our study to further isolate theimpact of financial sector development on the RE sector However it is stilltoo early to gauge the full impact of the new Kyoto financing possibilitieson the adoption of RE in developing and transition countries this questionis therefore left to future research5

The analysis of the role of the financial sector ndash commercial bankingfinancial markets insurance etc ndash for economic performance has generateda vast literature during the past two decades (see eg Demirguc-Kunt andLevine 2001 for an overview) The importance of the (private) financialsector particularly for the development of the energy sector has also beenpointed out in several studies In one of the earliest analyses of energysector financing in developing countries Churchill and Saunders (1989)discuss a proper policy framework to encourage private sector financialinvolvement Ten years later Babbar and Schuster (1998) and Head (2000)still find substantial gaps in the financing of power projects particularly REprojects The financing obstacles for RETs are confirmed in the overview byWohlgemuth and Painuly (1999) where efforts in different countries andregions are discussed and several policy recommendations derived Morerecently Sonntag-OrsquoBrien and Usher (2004a) and Painuly and Wohlgemuth(2006) take a look at the experience to date with RET implementation indeveloping and transition countries and again point to the (private sector)financing problems that RE projects encounter as well as to some successfulmodels that have been adopted MacLean and Siegel (2007) concentrate onthe financing of small-scale RE projects and distinguish three financingareas end-user finance business finance and small-scale project financeThe need for well-informed (local) commercial financiers often to act as

4 The Tejona wind farm also illustrates the importance of an adequate financingframework The projectrsquos history dates back to the 1970s when the Costa Ricangovernment applied to the regional Development Bank and the World Bank forfinancing assistance A feasibility study followed in 1993 and a private windenergy company sought to develop the wind farm as an independent powerproject However it took until 1998 to find a project financing model which endedup being a build-operate-lease construction together with a Dutch consortiumwith the support of the Dutch government and the coordination of the GlobalEnvironment Facility In the meantime the Costa Rican Electricity Institute haddecided to develop the site as a public sector project The wind farm is nowoperational and has a capacity of 20 MW Several private wind farm developersare negotiating project finance arrangements for new wind power developmentsunder power purchase agreements with the Electricity Institute (see van Hulleet al 2003)

5 See for example Huang and Barker (2009) for a study on the effects to date of theCDM mechanism on CO2 emissions

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 245

financing partners for the government andor international developmentorganizations is a recurrent theme in all three areas6

Despite the importance of the financial sector for the development of RETsparticularly in developing and transition countries which is borne out bynumerous case studies and anecdotal evidence the subject has receivedlittle academic research attention This paper contributes to the knowledgeon what determines RET implementation by empirically analysing therelationship between financial intermediation and RE sector developmentwith a focus on non-OECD countries To the best of our knowledge this isthe first attempt to examine this issue in a systematic empirical analysis7

The literature discussed above suggests that a more highly developedfinancial sector will have a positive impact on the development of theRE sector A well-functioning and unrestricted banking sector should beparticularly relevant for credit allocation to RE companies in developingand transition countries where financial markets are still relatively smalland bond or equity financing is therefore difficult or impossible Of coursea well-developed financial sector alone is no guarantee for the success ofRETs the availability of adequate financing mechanisms should be viewedin the wider context of an appropriate RE policy framework8

We propose an empirical framework to test the effect of financial sectordevelopment on RETs We construct a panel dataset for up to 119 non-OECDcountries for the period 1980ndash2006 using electricity generation per capita

6 Among the examples cited in the report is that of a small hydro project on the WestNile in Uganda The project was realized thanks to a loan from a commercial lenderBarclays which has a strong presence in Africa lsquoGiven their lack of experiencewith this type of project and other financial market conditionsrsquo Barclays washowever only willing to extend a seven-year loan (MacLean and Siegel 200771) In order to make the project affordable the World Bank provided a partialguarantee which extended the loan term to 14 years Note that since that initialexperience Barclays has financed similar projects in Africa including anothersmall hydro plant in Kenya and ndash together with a South African bank ndash a 250 MWhydro-electric station again on the Nile river in Uganda

7 The theoretical contributions on this topic are equally scarce one attempt tomodel the connections between finance and RE sector development is presentedin Brunnschweiler (2006)

8 Missing finance is obviously connected to the more general policy frameworkfor RETs as previous literature has pointed out limited financing of RETsderives not only from underdeveloped financial intermediators but also fromthe lack of a specific policy design andor crowding-out effects from governmentpolicies favouring investment in fossil fuel projects (Churchill and Saunders 1989Wohlgemuth and Painuly 1999 Head 2000 World Bank 2002 Sonntag-OrsquoBrienand Usher 2004b UNEP FI 2004) Institutional shortcomings also contribute to theoften limited consideration by potential investors of the positive environmentalexternalities of RETs in project development costs In general the perceptionthat energy sustainability is not a top priority for policymakers further lowersinvestorsrsquo willingness to finance projects where the foreseeable rewards are alreadyrelatively low and long in the coming (see Williams and Ghanadan 2006 for auseful survey of electricity reform policies in developing and transition countries)We take the policy framework into account in our estimations to isolate the specificeffect of financial sector development (see section 2)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

246 Christa N Brunnschweiler

from RE technologies as a proxy for RE sector development We isolate thefinancial sector effects by controlling for energy-relevant policy measuresas well as measures of more general institutional quality The empiricalresults from generalised least squares (GLS) and dynamic ArellanondashBondGeneralised Method-of-Moments (GMM) panel estimations confirm thepositive effect of financial sector and especially commercial banking sectordevelopment on RET use in developing and transition countries The effectsare particularly strong and economically significant for the newer RETsincluding wind solar geothermal and biomass while the overall impacton RE and hydropower generation is much more limited in magnitudethough still statistically significant

The results also suggest that the adoption of the Kyoto Protocol has hada strong positive effect on the diffusion of RETs throughout developingand transition countries The findings are robust to the inclusion of othercovariates which could influence RE sector development such as oilcoal and natural gas production and prices and proxies for RE potentialThe implication is that without proper (private sector) finance RE isunlikely to reach its full potential in the developing world An adequatefinancing framework should therefore be part of a more general RE sectordevelopment policy

The paper is organised as follows The data and empirical methodologyare described in section 2 while the results are presented in section 3 Section4 concludes with a brief summary and discussion

2 Data and methodologyThe discussion of the role of commercial finance in the development ofRE has so far been based mainly on case studies and anecdotal evidenceThe lack of a more systematic empirical analysis of the correlation betweenfinancial sector and RE development independent of (or complementaryto) a specific RE policy framework has probably also been due to the dataproblem regarding the quantification of the RE sector especially in thedeveloping world

The obstacles begin with the definition of RE in official statisticstraditionally hydropower ndash mostly provided by large plants ndash has deliveredthe lionrsquos share of RE in countriesrsquo energy generation mix with othertypes of RE ndash when included ndash making up barely a few per cent ofoverall energy production Recently however some environmentalists andpolicymakers have contended that large hydropower projects should notbe viewed as viable contributions to sustainable energy production asthey often cause serious negative environmental and social externalities(notable examples are the giant Three Gorges Dam on the Yangtze Riverin China or the Ilisu dam project in Turkey) Moreover most traditionallarge hydro projects in the developing world have been co-financed bymultilateral financial institutions and the local governments with little orno involvement sought of commercial finance (World Bank 2003) Theuse of an overall RE measure could therefore introduce a downwardbias into the results on the importance of the financial sector for moremodern RETs such as wind geothermal and solar power A further possibleissue concerns the negative impact on agricultural (food) production of

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 4: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

244 Christa N Brunnschweiler

emission-reduction projects in non-Annex B developing countries andearn certified emission reduction credits (JI programmes on the other handinvolve joint projects by two Annex B countries) Some RE projects werealready initiated in the late 1990s as CDM or JI pilot programmes suchas the Tejona wind farm in Costa Rica4 The Kyoto Protocolrsquos adoption inDecember 1997 obviously marks a major change in global climate policy weincorporate the potential lsquoKyoto effectsrsquo into our study to further isolate theimpact of financial sector development on the RE sector However it is stilltoo early to gauge the full impact of the new Kyoto financing possibilitieson the adoption of RE in developing and transition countries this questionis therefore left to future research5

The analysis of the role of the financial sector ndash commercial bankingfinancial markets insurance etc ndash for economic performance has generateda vast literature during the past two decades (see eg Demirguc-Kunt andLevine 2001 for an overview) The importance of the (private) financialsector particularly for the development of the energy sector has also beenpointed out in several studies In one of the earliest analyses of energysector financing in developing countries Churchill and Saunders (1989)discuss a proper policy framework to encourage private sector financialinvolvement Ten years later Babbar and Schuster (1998) and Head (2000)still find substantial gaps in the financing of power projects particularly REprojects The financing obstacles for RETs are confirmed in the overview byWohlgemuth and Painuly (1999) where efforts in different countries andregions are discussed and several policy recommendations derived Morerecently Sonntag-OrsquoBrien and Usher (2004a) and Painuly and Wohlgemuth(2006) take a look at the experience to date with RET implementation indeveloping and transition countries and again point to the (private sector)financing problems that RE projects encounter as well as to some successfulmodels that have been adopted MacLean and Siegel (2007) concentrate onthe financing of small-scale RE projects and distinguish three financingareas end-user finance business finance and small-scale project financeThe need for well-informed (local) commercial financiers often to act as

4 The Tejona wind farm also illustrates the importance of an adequate financingframework The projectrsquos history dates back to the 1970s when the Costa Ricangovernment applied to the regional Development Bank and the World Bank forfinancing assistance A feasibility study followed in 1993 and a private windenergy company sought to develop the wind farm as an independent powerproject However it took until 1998 to find a project financing model which endedup being a build-operate-lease construction together with a Dutch consortiumwith the support of the Dutch government and the coordination of the GlobalEnvironment Facility In the meantime the Costa Rican Electricity Institute haddecided to develop the site as a public sector project The wind farm is nowoperational and has a capacity of 20 MW Several private wind farm developersare negotiating project finance arrangements for new wind power developmentsunder power purchase agreements with the Electricity Institute (see van Hulleet al 2003)

5 See for example Huang and Barker (2009) for a study on the effects to date of theCDM mechanism on CO2 emissions

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 245

financing partners for the government andor international developmentorganizations is a recurrent theme in all three areas6

Despite the importance of the financial sector for the development of RETsparticularly in developing and transition countries which is borne out bynumerous case studies and anecdotal evidence the subject has receivedlittle academic research attention This paper contributes to the knowledgeon what determines RET implementation by empirically analysing therelationship between financial intermediation and RE sector developmentwith a focus on non-OECD countries To the best of our knowledge this isthe first attempt to examine this issue in a systematic empirical analysis7

The literature discussed above suggests that a more highly developedfinancial sector will have a positive impact on the development of theRE sector A well-functioning and unrestricted banking sector should beparticularly relevant for credit allocation to RE companies in developingand transition countries where financial markets are still relatively smalland bond or equity financing is therefore difficult or impossible Of coursea well-developed financial sector alone is no guarantee for the success ofRETs the availability of adequate financing mechanisms should be viewedin the wider context of an appropriate RE policy framework8

We propose an empirical framework to test the effect of financial sectordevelopment on RETs We construct a panel dataset for up to 119 non-OECDcountries for the period 1980ndash2006 using electricity generation per capita

6 Among the examples cited in the report is that of a small hydro project on the WestNile in Uganda The project was realized thanks to a loan from a commercial lenderBarclays which has a strong presence in Africa lsquoGiven their lack of experiencewith this type of project and other financial market conditionsrsquo Barclays washowever only willing to extend a seven-year loan (MacLean and Siegel 200771) In order to make the project affordable the World Bank provided a partialguarantee which extended the loan term to 14 years Note that since that initialexperience Barclays has financed similar projects in Africa including anothersmall hydro plant in Kenya and ndash together with a South African bank ndash a 250 MWhydro-electric station again on the Nile river in Uganda

7 The theoretical contributions on this topic are equally scarce one attempt tomodel the connections between finance and RE sector development is presentedin Brunnschweiler (2006)

8 Missing finance is obviously connected to the more general policy frameworkfor RETs as previous literature has pointed out limited financing of RETsderives not only from underdeveloped financial intermediators but also fromthe lack of a specific policy design andor crowding-out effects from governmentpolicies favouring investment in fossil fuel projects (Churchill and Saunders 1989Wohlgemuth and Painuly 1999 Head 2000 World Bank 2002 Sonntag-OrsquoBrienand Usher 2004b UNEP FI 2004) Institutional shortcomings also contribute to theoften limited consideration by potential investors of the positive environmentalexternalities of RETs in project development costs In general the perceptionthat energy sustainability is not a top priority for policymakers further lowersinvestorsrsquo willingness to finance projects where the foreseeable rewards are alreadyrelatively low and long in the coming (see Williams and Ghanadan 2006 for auseful survey of electricity reform policies in developing and transition countries)We take the policy framework into account in our estimations to isolate the specificeffect of financial sector development (see section 2)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

246 Christa N Brunnschweiler

from RE technologies as a proxy for RE sector development We isolate thefinancial sector effects by controlling for energy-relevant policy measuresas well as measures of more general institutional quality The empiricalresults from generalised least squares (GLS) and dynamic ArellanondashBondGeneralised Method-of-Moments (GMM) panel estimations confirm thepositive effect of financial sector and especially commercial banking sectordevelopment on RET use in developing and transition countries The effectsare particularly strong and economically significant for the newer RETsincluding wind solar geothermal and biomass while the overall impacton RE and hydropower generation is much more limited in magnitudethough still statistically significant

The results also suggest that the adoption of the Kyoto Protocol has hada strong positive effect on the diffusion of RETs throughout developingand transition countries The findings are robust to the inclusion of othercovariates which could influence RE sector development such as oilcoal and natural gas production and prices and proxies for RE potentialThe implication is that without proper (private sector) finance RE isunlikely to reach its full potential in the developing world An adequatefinancing framework should therefore be part of a more general RE sectordevelopment policy

The paper is organised as follows The data and empirical methodologyare described in section 2 while the results are presented in section 3 Section4 concludes with a brief summary and discussion

2 Data and methodologyThe discussion of the role of commercial finance in the development ofRE has so far been based mainly on case studies and anecdotal evidenceThe lack of a more systematic empirical analysis of the correlation betweenfinancial sector and RE development independent of (or complementaryto) a specific RE policy framework has probably also been due to the dataproblem regarding the quantification of the RE sector especially in thedeveloping world

The obstacles begin with the definition of RE in official statisticstraditionally hydropower ndash mostly provided by large plants ndash has deliveredthe lionrsquos share of RE in countriesrsquo energy generation mix with othertypes of RE ndash when included ndash making up barely a few per cent ofoverall energy production Recently however some environmentalists andpolicymakers have contended that large hydropower projects should notbe viewed as viable contributions to sustainable energy production asthey often cause serious negative environmental and social externalities(notable examples are the giant Three Gorges Dam on the Yangtze Riverin China or the Ilisu dam project in Turkey) Moreover most traditionallarge hydro projects in the developing world have been co-financed bymultilateral financial institutions and the local governments with little orno involvement sought of commercial finance (World Bank 2003) Theuse of an overall RE measure could therefore introduce a downwardbias into the results on the importance of the financial sector for moremodern RETs such as wind geothermal and solar power A further possibleissue concerns the negative impact on agricultural (food) production of

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 5: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 245

financing partners for the government andor international developmentorganizations is a recurrent theme in all three areas6

Despite the importance of the financial sector for the development of RETsparticularly in developing and transition countries which is borne out bynumerous case studies and anecdotal evidence the subject has receivedlittle academic research attention This paper contributes to the knowledgeon what determines RET implementation by empirically analysing therelationship between financial intermediation and RE sector developmentwith a focus on non-OECD countries To the best of our knowledge this isthe first attempt to examine this issue in a systematic empirical analysis7

The literature discussed above suggests that a more highly developedfinancial sector will have a positive impact on the development of theRE sector A well-functioning and unrestricted banking sector should beparticularly relevant for credit allocation to RE companies in developingand transition countries where financial markets are still relatively smalland bond or equity financing is therefore difficult or impossible Of coursea well-developed financial sector alone is no guarantee for the success ofRETs the availability of adequate financing mechanisms should be viewedin the wider context of an appropriate RE policy framework8

We propose an empirical framework to test the effect of financial sectordevelopment on RETs We construct a panel dataset for up to 119 non-OECDcountries for the period 1980ndash2006 using electricity generation per capita

6 Among the examples cited in the report is that of a small hydro project on the WestNile in Uganda The project was realized thanks to a loan from a commercial lenderBarclays which has a strong presence in Africa lsquoGiven their lack of experiencewith this type of project and other financial market conditionsrsquo Barclays washowever only willing to extend a seven-year loan (MacLean and Siegel 200771) In order to make the project affordable the World Bank provided a partialguarantee which extended the loan term to 14 years Note that since that initialexperience Barclays has financed similar projects in Africa including anothersmall hydro plant in Kenya and ndash together with a South African bank ndash a 250 MWhydro-electric station again on the Nile river in Uganda

7 The theoretical contributions on this topic are equally scarce one attempt tomodel the connections between finance and RE sector development is presentedin Brunnschweiler (2006)

8 Missing finance is obviously connected to the more general policy frameworkfor RETs as previous literature has pointed out limited financing of RETsderives not only from underdeveloped financial intermediators but also fromthe lack of a specific policy design andor crowding-out effects from governmentpolicies favouring investment in fossil fuel projects (Churchill and Saunders 1989Wohlgemuth and Painuly 1999 Head 2000 World Bank 2002 Sonntag-OrsquoBrienand Usher 2004b UNEP FI 2004) Institutional shortcomings also contribute to theoften limited consideration by potential investors of the positive environmentalexternalities of RETs in project development costs In general the perceptionthat energy sustainability is not a top priority for policymakers further lowersinvestorsrsquo willingness to finance projects where the foreseeable rewards are alreadyrelatively low and long in the coming (see Williams and Ghanadan 2006 for auseful survey of electricity reform policies in developing and transition countries)We take the policy framework into account in our estimations to isolate the specificeffect of financial sector development (see section 2)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

246 Christa N Brunnschweiler

from RE technologies as a proxy for RE sector development We isolate thefinancial sector effects by controlling for energy-relevant policy measuresas well as measures of more general institutional quality The empiricalresults from generalised least squares (GLS) and dynamic ArellanondashBondGeneralised Method-of-Moments (GMM) panel estimations confirm thepositive effect of financial sector and especially commercial banking sectordevelopment on RET use in developing and transition countries The effectsare particularly strong and economically significant for the newer RETsincluding wind solar geothermal and biomass while the overall impacton RE and hydropower generation is much more limited in magnitudethough still statistically significant

The results also suggest that the adoption of the Kyoto Protocol has hada strong positive effect on the diffusion of RETs throughout developingand transition countries The findings are robust to the inclusion of othercovariates which could influence RE sector development such as oilcoal and natural gas production and prices and proxies for RE potentialThe implication is that without proper (private sector) finance RE isunlikely to reach its full potential in the developing world An adequatefinancing framework should therefore be part of a more general RE sectordevelopment policy

The paper is organised as follows The data and empirical methodologyare described in section 2 while the results are presented in section 3 Section4 concludes with a brief summary and discussion

2 Data and methodologyThe discussion of the role of commercial finance in the development ofRE has so far been based mainly on case studies and anecdotal evidenceThe lack of a more systematic empirical analysis of the correlation betweenfinancial sector and RE development independent of (or complementaryto) a specific RE policy framework has probably also been due to the dataproblem regarding the quantification of the RE sector especially in thedeveloping world

The obstacles begin with the definition of RE in official statisticstraditionally hydropower ndash mostly provided by large plants ndash has deliveredthe lionrsquos share of RE in countriesrsquo energy generation mix with othertypes of RE ndash when included ndash making up barely a few per cent ofoverall energy production Recently however some environmentalists andpolicymakers have contended that large hydropower projects should notbe viewed as viable contributions to sustainable energy production asthey often cause serious negative environmental and social externalities(notable examples are the giant Three Gorges Dam on the Yangtze Riverin China or the Ilisu dam project in Turkey) Moreover most traditionallarge hydro projects in the developing world have been co-financed bymultilateral financial institutions and the local governments with little orno involvement sought of commercial finance (World Bank 2003) Theuse of an overall RE measure could therefore introduce a downwardbias into the results on the importance of the financial sector for moremodern RETs such as wind geothermal and solar power A further possibleissue concerns the negative impact on agricultural (food) production of

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 6: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

246 Christa N Brunnschweiler

from RE technologies as a proxy for RE sector development We isolate thefinancial sector effects by controlling for energy-relevant policy measuresas well as measures of more general institutional quality The empiricalresults from generalised least squares (GLS) and dynamic ArellanondashBondGeneralised Method-of-Moments (GMM) panel estimations confirm thepositive effect of financial sector and especially commercial banking sectordevelopment on RET use in developing and transition countries The effectsare particularly strong and economically significant for the newer RETsincluding wind solar geothermal and biomass while the overall impacton RE and hydropower generation is much more limited in magnitudethough still statistically significant

The results also suggest that the adoption of the Kyoto Protocol has hada strong positive effect on the diffusion of RETs throughout developingand transition countries The findings are robust to the inclusion of othercovariates which could influence RE sector development such as oilcoal and natural gas production and prices and proxies for RE potentialThe implication is that without proper (private sector) finance RE isunlikely to reach its full potential in the developing world An adequatefinancing framework should therefore be part of a more general RE sectordevelopment policy

The paper is organised as follows The data and empirical methodologyare described in section 2 while the results are presented in section 3 Section4 concludes with a brief summary and discussion

2 Data and methodologyThe discussion of the role of commercial finance in the development ofRE has so far been based mainly on case studies and anecdotal evidenceThe lack of a more systematic empirical analysis of the correlation betweenfinancial sector and RE development independent of (or complementaryto) a specific RE policy framework has probably also been due to the dataproblem regarding the quantification of the RE sector especially in thedeveloping world

The obstacles begin with the definition of RE in official statisticstraditionally hydropower ndash mostly provided by large plants ndash has deliveredthe lionrsquos share of RE in countriesrsquo energy generation mix with othertypes of RE ndash when included ndash making up barely a few per cent ofoverall energy production Recently however some environmentalists andpolicymakers have contended that large hydropower projects should notbe viewed as viable contributions to sustainable energy production asthey often cause serious negative environmental and social externalities(notable examples are the giant Three Gorges Dam on the Yangtze Riverin China or the Ilisu dam project in Turkey) Moreover most traditionallarge hydro projects in the developing world have been co-financed bymultilateral financial institutions and the local governments with little orno involvement sought of commercial finance (World Bank 2003) Theuse of an overall RE measure could therefore introduce a downwardbias into the results on the importance of the financial sector for moremodern RETs such as wind geothermal and solar power A further possibleissue concerns the negative impact on agricultural (food) production of

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 7: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 247

Table 1 Descriptive statistics

Shown for non-OECD countries from 1980 to 2006

Variable Observation MeanStandarddeviation Minimum Maximum

repc 3911 031 078 000 1003hydropc 3911 030 078 000 1003geopc 3911 001 002 000 034dbacba 3321 074 023 minus011 134pcrdbgdp 2901 027 023 000 166llgdp 2914 041 027 000 157psreform 3564 258 237 000 600econfree 2446 55 11 21 879gdppc 3566 494610 706455 11176 7071584fdigdp 3380 330 922 minus8289 34819oilprodpc 3967 004 017 000 211oilprice 4212 2427 1131 1221 6150hydropot 3022 9325 26213 0 2474windpot 4212 013 034 0 1

encouraging biomass production for use as biofuel as demonstrated by therecent large fluctuations in the prices of grains and other foodstuffs

We consider these issues when testing the importance of financialintermediation for RET development First we distinguish betweendifferent types of RE generation in addition to the aggregate measureSecond as a proxy for our dependent variable RE sector development weuse electricity generated with renewable resources in per-capita terms andtherefore avoid the issue of biofuels

We construct three separate measures of RE sector development Thefirst repc measures the overall renewable resource electricity generation ndashincluding all types of hydropower wood and waste geothermal solar andwind ndash in billion kwh per capita The second variable hydropc considersonly hydroelectric power generation again in billion kwh per capitaThis distinction takes into account the importance of large hydropowerin electricity generation and their possible distorting effect on the resultsfound using the data on total RE generation The third and final dependentvariable geopc considers electricity produced from all non-hydro RE typesincluding some of the latest RETs ie geothermal solar wind and woodand waste energy resources Again this measure is in billion kwh per capita

The electricity generation data for all three dependent variables is freelyavailable from the US Energy Information Administration (EIA) on a yearlybasis since 1980 (or since the early 1990s for countries of the former SovietUnion and former Yugoslavia) The EIA notes that the sum of componentsin the data may not equal the total listed due to independent roundingDetailed descriptions of all variables and their sources can be found inthe appendix The descriptive statistics in table 1 show that the dependent

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 8: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

248 Christa N Brunnschweiler

variables repc and hydropc have a wide variation in per capita RE resourceintensity while the generation of non-hydro renewable resource electricity(geopc) is more limited

A look at the share of RE in overall electricity generation shows that onaverage in non-OECD countries around 34 per cent of electric power wasproduced by RE resources between 1980 and 2006 compared with around 32per cent in OECD countries This relatively similar picture remains when weconsider the share of hydroelectric power generation in overall electricitygeneration which was on average around 33 per cent for non-OECD and29 per cent for OECD countries In both groups of countries hydropowercovered from 0 per cent right up to 100 per cent of total electricity generationThe situation is more varied when we look at the non-hydro RE share innon-OECD countries wood waste wind solar and geothermal energyproduced barely 1 per cent of total electricity on average but up to 40 percent in low- to middle-income countries like El Salvador and Nicaragua InOECD countries the average share of electricity generated from non-hydroRE was three per cent reaching up to nearly 20 per cent in Luxembourgand 30 per cent in Denmark9

As noted above the adoption of the Kyoto Protocol in late 1997 marksa huge shift in global climate policy which also affects RE policy It istherefore interesting to examine the data for any prima facie evidence of achange in RE use since 1998 In general there has indeed been an increasein electricity generation with RETs in non-OECD countries overall REelectricity generation went from an average of 2634 million kwh per capitabefore Kyoto to 3938 million kwh per capita post-Kyoto This was due toa large increase in hydropower but also to a near doubling of non-hydroelectricity production from an average 5 million kwh per capita to 96 millionkwh per capita Further analysis reveals that the post-Kyoto increase in REuse is common for all country income groups It is also observable in allregions of the world with a particularly large jump in RE use in South-Eastern Asia here average overall RE electricity generation went from2958 billion kwh per capita to 5078 billion kwh per capita post-KyotoAgain most of the increase in this region comes from more hydropowerbut non-hydro electricity generation increased more than 27-fold from acomparatively very modest 21 thousand kwh per capita to an average 5694thousand kwh per capita in the period after 1998 These initial findings aresuggestive of a strong lsquoKyoto effectrsquo which we will test for in the estimationsbelow

9 In our estimations we use the per-capita values of RE and not the shareof RE in total energy production This avoids picking up possible spuriouscorrelations between financial sector development and the RE share since totalenergy generation (in the denominator) is probably highly correlated with overalleconomic development (including financial sector development) To furtherpreclude spurious results we control for income per capita (see below) Howeverthe main results ndash in particular for the share of non-hydro RE used in powergeneration ndash are qualitatively similar using RE shares to those found using RE percapita (available upon request)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 9: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 249

The main explanatory variables include three different indicators offinancial sector development and a vector of control variables Themeasures of financial sector development are taken from Beck et al (2000)They are not direct measures of banksrsquo efficiency in credit allocation butrather different proxies for financial intermediary development tested inthe literature10 The first indicator of financial sector development dbacbameasures the importance of commercial banksrsquo asset share versus thatof the central bank In more highly developed and open economies thecommercial financial sector handles a greater share of household savingsthan the central bank Assuming that the commercial financial sector is moreefficient than the public one in allocating credits dbacba should positivelycorrelate with RET development This variable has also been tested severaltimes in the finance literature eg in King and Levine (1993ab) and Levineet al (2000)

The second variable pcrdbgdp captures the amount of credit providedby financial institutions to the private sector as a share of gross domesticproduct (GDP) It excludes credits issued by governments and developmentbanks An unrestricted financial sector can be expected to account fora larger share of lending to the private sector In fact this variable hasbeen shown by Levine et al (2000) to be a reliable measure of financialintermediary development ie the ability of financial institutions toefficiently mobilise and allocate resources to profitable ventures Earlierversions of the measure were used for example in King and Levine (1993ab)and Levine and Zervos (1998) We expect pcrdbgdp to correlate positivelywith the level of development of the RE sector

The third and final financial variable llgdp is a general measure offinancial sector development commonly known as lsquofinancial depthrsquo It isdefined as liquid liabilities of the financial system ndash currency plus demandand interest-bearing liabilities of banks and other financial intermediariesor more generally M2 ndash divided by GDP Financial depth is the broadestmeasure of financial intermediation giving an indication of the overall sizeof the financial sector without distinguishing either between commercialand non-commercial banks and other financial intermediaries or betweenthe use of the liabilities The assumption is that the relative size of thefinancial intermediary sector is positively correlated with the quantity andquality of the financial services provided and we would therefore againexpect a positive influence on the development of RETs

All three financial variables are measured in current prices and GDPis purchasing power parity (PPP) adjusted (see Beck et al 2000 for more

10 Beck et al (2000) provide a large financial structure dataset on the World Bankwebsite The data used here comes from the dataset revised on November 212008 This also includes two new variables which measure the efficiency withwhich commercial banks channel funds from savers to investors overhead costs(ie the accounting value of a bankrsquos overhead costs as share of its total assets) andthe net interest margin (ie the accounting value of a bankrsquos net interest revenueas a share of its total assets) Unfortunately however these measures are as yetonly available since the mid 1990s for a limited number of countries We thereforeuse three more conventional measures of financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 10: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

250 Christa N Brunnschweiler

details) Table 1 shows that there is considerable variation in the financialsector development in non-OECD countries ranging from practicallynon-existent in some countries to levels comparable with many OECDcountries11 As is to be expected the means of all three indices were higherthroughout the period for upper-middle-income and high-income non-OECD countries the average values of dbacba pcrdbgdp and llgdp wererespectively 086 039 and 053 as opposed to 067 019 and 033 in low-and lower-middle-income countries These numbers also seem to indicatethat the three variables measure slightly different aspects of the financialsector in fact dbacba has a 05 correlation with pcrdbgdp and 039 withllgdp while pcrdbgdp and llgdp are more highly correlated (08) We will usethe three financial sector variables separately to minimise multicollinearityissues

We expect RE sector development to depend on several factors other thanfinancial intermediation One obvious factor is the regulatory (or policy)framework for RE We introduce two variables to capture different aspectsof energy sector regulation The first psreform describes the level of powersector reform without special reference to RETs It is based on a broadqualitative survey by the World Bank conducted in 1998 (ESMAP 1999) andtakes on values from 0 (least reformed) to 6 (reforms in all relevant areashave been implemented) The evaluation considers measures to create equalmarket opportunities for all energy resource types and encourage privatefirmsrsquo participation and competition (lsquocompetitionrsquo being a main reformcriterion) Hence psreform is a proxy for government energy policies ndashalthough unfortunately a time-invariant proxy since it is based on a one-time study The ESMAP study was conducted for a large number ofdeveloping and transition economies but not for most high-income andOECD countries We assigned a value of 6 to all high-income non-OECDeconomies ndash ten mainly small island states ndash and the OECD countries (usedin comparison estimations) In robustness tests we add a dummy variablefor lsquoartificiallyrsquo assigned top scores or alternatively drop these countriesaltogether both methods do not alter our main results and even reinforcethem (see below) The descriptive statistics show that power sector reformshave a relatively low mean of 258

The second policy measure seeks to capture more specific RE policiesby looking at the effects of the adoption of the Kyoto Protocol which alsoincludes efforts to diversify energy production to include more RETs Forthis we construct a simple zero-one dummy variable that takes on valueone for all years from 1998 onwards The adoption of the Kyoto Protocolarguably also marks a greater awareness of environmental issues whichis not limited to industrialised countries The level of lsquoenvironmentalismrsquomay also contribute to the diffusion of RETs

As discussed in the previous section the institutional frameworkis also a crucial element of financiersrsquo information costs on RETssignalling a governmentrsquos commitment to levelling the playing field for

11 Note that the most highly developed OECD countries have a financial depth (llgdp)of around 37 a commercial bank asset share (dbacba) of 1 and a private creditshare (pcrdbgdp) of nearly 28

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 11: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 251

energy providers or even positively encouraging RETs thereby reducinguncertainty about future profitability of a RE project Since there is noreliable data available on creditor evaluation costs in less developedeconomies these crude policy proxies will have to suffice We expect apositive impact of both power sector reforms and the Kyoto dummy on theRE sector

In addition to the two energy policy variables we also use measuresof overall institutional quality in robustness tests It is in fact likely thatRE projects like other types of investment projects benefit from generalpolitical stability sound regulatory frameworks effective governance andsecure property rights Moreover institutional quality in general couldbe correlated with financial sector development in particular althoughcorrelation coefficients range between a modest 038 and 05 (with theexception of pcrdbgdp which has correlation coefficients between 052 and061 with the institutional quality measures) We consider three differentinstitutional measures the first is an economic freedom index compiled byGwartney et al (2008) on a five yearly basis until 2000 and on a yearly basissince 2001 (values for the most recent year were used for intermediateyears before 2000) The index ranges from 1 (worst) to 10 (best) andincludes evaluations of the legal structure and security of property rightsgovernment size access to sound money trade freedom and regulation ofcredit labour and business The other two measures capture regulatoryquality and government effectiveness respectively and are taken fromKaufmann et al (2008) These are available for a wider range of countriesthan the economic freedom index but they have only been compiled since1996 Although institutions change only slowly we concentrate on theeconomic freedom index and only briefly discuss results using the othertwo measures where the 1996 values were used for earlier years

Several other control variables are included Income per capita in US$(gdppc) controls for the possibility that richer and economically moredeveloped countries may simply have higher energy production The ratioof net foreign direct investment inflows to GDP (fdigdp) accounts for non-domestic investment including investment by foreign development banksWe would generally expect this measure to positively affect RE sectordevelopment12

The prices of the most common conventional fuels may affect investmentin alternative energy sources We therefore control for the possibleexogenous effects on RE development of the costs of non-RE resourcesproduction by including the average annual market price of crude oil(oilprice) coal (coalprice) and natural gas (natgasprice) as well as therespective per capita production rates (oilprodpc coalprodpc and natgaspdopc)Furthermore an interaction term controls for the possibility that large fossilfuel producers react differently to price changes than small producers or

12 In further tests we also consider official development assistance by multilateralsas well as a variation including aid to control for the specific effect of multilateraldonor money This variable proved insignificant and is therefore not shownDummy variables for income groups (classified according to the World Bank)were also not robust and did not change the main results

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 12: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

252 Christa N Brunnschweiler

countries without any fossil fuels at all In fossil fuel poor countries wewould expect a clear positive effect of a fossil fuel price increase on theshare of RE in power production Conversely the more fossil fuels a countryproduces the less likely it will be to invest in alternative energy resourcesparticularly in times of high prices and returns

We also control for regional effects as well as two specific time effectsin the estimation period namely the fall of the Soviet bloc and the startof the economic (and political) transition of Central and Eastern Europeanand Central Asian countries and the financial crisis of East Asia LatinAmerica and Russia We introduce a variable called transition for the formerevent which has value zero until 1991 and then increases with every yearafter transition for transition countries remaining zero for all others Thisis meant to account for the time-diminishing overall negative economiceffect of the transition shock The variable fincrisis captures the secondevent (common to all countries) assuming value zero until 1997 and thenincreasing with every successive post-crisis year Both time variables areexpected to have a positive effect on RE sector development as investmentin developing and transition countries generally slumped after the shocksand then gradually picked up again

Finally we seek to control for RE potential by including two variablesfor the two most important and widely diffused RETs hydropot measuresthe technically exploitable part of gross theoretical hydropower capabilityin twh per year Where not available gross theoretical capability was usedinstead The data was taken from the World Energy Council triannual Surveyof Energy Resources (WEC 1980ndash2007) intermediate years were filled in withthe latest available value windpot measures the potential for the currentlymost widely employed non-hydro RET namely wind power The measureis based on a recent study by Archer and Jacobson (2005) who present thefirst attempt at quantifying global wind power potential from real data Weconstruct a dummy variable where zero equals poor or very poor potentialand one equals moderate to high potential According to the authors ofthe study moderate (economic) potential starts at wind speeds at 80 m(the hub-height of modern 77 m diameter 1500 kw turbines) between 68and 75 ms They find substantial potential in all regions particularly inNorth America and (northern) Europe as well as the southern tip of SouthAmerica Interestingly the data of Archer and Jacobson (2005) show thatno wind-speed reporting station in mainland China suggests moderate tohigh wind power potential This runs counter to other studies focusing onChina and to the increasing number of wind farms in China itself Althoughtheir study is the first to calculate wind power potential with a consistentmethodology for the entire world their data is available for one year only(generally 2000) which may be problematic if wind patterns change overtime for example due to effects of El Nino

The EIA RE dataset contains several missing years due to newlyindependent countries and gives us an unbalanced panel covering 119 non-OECD countries for the period from 1980 to 200613 Results including OECDcountries are also shown for comparison

13 Missing years in the dependent variables were not changed Missing years inthe explanatory variables appear to be random up to two missing years were

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 13: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 253

In our main estimations we perform GLS regressions for the equation

Yit = α1 + α2 Fit + α3 Xit + ωi t (1)

where Yit is the dependent variable (repc hydropc or geopc) in country i attime t Fit denotes the financial sector development variable and Xit thevector of control variables The composite error term ωit consists of thecountry-specific error component εi and the combined cross-section andtime series error component uit according to ωit = εi + uit14

The main estimations were performed with one-year-lags for all relevantindependent variables since financial sector and other economic changesare not expected to have immediate effects on electricity generationAlternative lag durations of up to five years confirm the results (availableupon request) but reduce the number of observations As there is no cleartheoretical indication regarding the ideal lag number we show resultsusing one-year-lags15 Time-invariant measures ndash the Kyoto transition andfinancial crisis variables ndash were not lagged

In further sensitivity tests that account for the possible dynamic effectsof RE sector development and test the causality we perform a dynamicpanel data analysis Linear dynamic panel analysis following Arellano andBond (1991) is also designed for panels where the cross-section dimensionexceeds the time dimension as in our case where we have up to 119 non-OECD countries and a maximum of 26 years Dynamic models includeunobserved country-level effects which by construction are correlated withthe lagged dependent variable making standard estimators inconsistentThe Arellano and Bond (1991) first-differenced GMM estimator is consistentfor the parameters of this model though it still requires that there be nosecond-order serial autocorrelation in the idiosyncratic errors which wesystematically test for after two-step GMM estimations (not shown)

The basic estimation equation remains much the same as above(excluding time-invariant regressors) with the addition of the laggeddependent variable Yi(t minus 1)

Yit = β0Yi(tminus1) + β1 + β2 Fit + β3 Xit + νi t (2)

completed with simple linear interpolation larger holes in the data were leftunaltered Dummy variables for interpolated years were insignificant

14 See for example Baltagi (2008) for an extensive discussion of panel data analysismodels

15 A further point worth considering is that financial sector development ndash ourmain variable of interest ndash changes only slowly over time which is an argumentin favour of using one-year-lags to maximise the number of observations Totest for possible cyclical effects estimations were also performed with five-yearaverage values for the dependent variables using beginning-of-period values forthe independent variables The estimations consistently showed positive effectsof financial intermediation on RE sector development but the impact was seldomsignificant Moreover the reduced number of observations generally led to lowstatistical quality of the estimations

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 14: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

254 Christa N Brunnschweiler

3 Estimation resultsIt is of particular interest to observe the sign and statistical validity of thefinancial sector coefficients α2 and β2 The aim is to observe whether thedevelopment of the RE sector is ndash other things equal ndash positively influencedby the financial intermediary sector in general (captured by the broadfinancial depth measure llgdp) and especially by the commercial bankingsystem (proxied by the commercial bank asset share dbacba and the privatecredit allocation pcrdbgdp) Controlling for variables which affect RE orfinancial sector development allows us to draw conclusions on the causalimpact of finance on RETs

31 Generalised least squares estimationsWe begin by presenting random-effects GLS estimations according toequation (1) for our unbalanced panel of non-OECD countries16 Table 2shows results for a parsimonious basic specification including a financialsector variable income per capita FDIGDP the two energy sector policymeasures psreform and kyoto and regional dummy variables Columns (1)ndash(3) give results with the total RE produced per capita (repc) while columns(4)ndash(6) show results for hydroelectric power (hydropc) and columns (7)ndash(9)for non-hydro power generation (geopc)

First of all it is striking that all three financial sector developmentmeasures have the expected positive sign in all but the very lastspecification and are moreover mostly significant As far as the magnitudeof the effects is concerned in terms of beta coefficients we see for examplefrom column (1) that a one-standard-deviation increase in the commercialbank asset share would ndash other things equal ndash lead to a small increase of003 standard deviations in the RE per capita produced ((023 times 0091)078)The effects on the overall RE generation per capita are slightly smaller inmagnitude for the other two measures of financial sector developmentweighing in at around 002 The effects are similarly minimal when weconsider only hydro power with beta coefficients again around 002However the effect of financial intermediation appears to be much moresubstantial when we consider only non-hydro RE a one-standard-deviationincrease in the commercial bank asset share leads to an increase in geopcby 0184 of a standard deviation (using results from column (7)) Thelarge negative effect of lsquofinancial depthrsquo in the last column is puzzlingit appears that financial sector development in general diminishes RET usewhile commercial banking increases it However this effect is not robust todropping outliers (see below)

Regarding the other covariates in table 2 we find on the one hand thatpower sector reforms have a counter-intuitive effect on RE generation thecoefficient has a negative sign in all but the estimations using non-hydro

16 Hausman specification tests consistently showed no advantage of using fixed-effects estimations and we therefore show only random-effects results to includethe impact of time-invariant variables Results with fixed-effects estimations(available upon request) were very similar to the random-effects estimations buthad lower explanatory power See Baltagi (2008) for more details on the Hausmanspecification test

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 15: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environm

entandD

evelopmentE

conomics

255Table 2 Financial development and per capita RE generation in non-OECD countries

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0091c 0073c 0016c

(363) (298) (583)pcrdbgdp 0082b 0079b 0002

(231) (228) (059)llgdp 0053 0061a minus0008b

(145) (171) (208)fdigdp 0001c minus00003 minus00002 0001c minus00002 minus00002 minus000001 minus00001 minus00001

(346) (036) (034) (354) (026) (023) (024) (092) (098)gdppc 000001c 0000003 0000003a 0000004b 0000002 0000002 0000001c 0000001c 0000001c

(303) (158) (187) (236) (103) (123) (454) (384) (480)psreform minus0099b minus0102b minus0102b minus01b minus0103b minus0104b 0001 0001 0001

(233) (220) (221) (234) (223) (224) (079) (117) (132)kyoto 0026c 0038c 0038c 0023c 0033c 0032c 0004c 0006c 0006c

(367) (485) (480) (327) (425) (415) (519) (651) (685)eca 0217 0278 0272 0229 0293 0289 minus0014c minus0016b minus0017c

(092) (105) (103) (097) (111) (109) (261) (256) (279)mena minus0706b minus0727a minus0732a minus0694b minus0714a minus0722a minus0013a minus0013 minus0011

(210) (190) (192) (207) (187) (190) (167) (153) (122)ssa minus0655c minus0670c minus0672c minus0648c minus0660c minus0662c minus0009a minus0011b minus0012b

(299) (287) (288) (296) (283) (284) (176) (204) (216)Observations 2450 2179 2192 2450 2179 2192 2450 2179 2192Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 003 003 002 002 007 005 005R2 between 012 012 012 012 012 012 008 013 013R2 overall 010 011 010 010 010 010 009 010 011

Notes All regressions are random-effects GLS on sample panel of non-OECD countries from 1980 to 2006 with one-year-lags for allindicators except psreform kyoto and the three regional dummies Absolute z-statistics in parentheses Constant term included in allspecifications (not shown) a b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 16: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

256 Christa N Brunnschweiler

RE Note however that in robustness tests including a dummy variable forcountries that were lsquoartificiallyrsquo assigned a top score of six for power sectorreforms or alternatively dropped from the estimations the significance onthe negative coefficients disappeared while the positive coefficients becamesignificant (results available upon request) On the other hand the resultsshow a consistently strong positive RE sector development trend since theadoption of the Kyoto Protocol suggesting that the connected emission-reduction policies (or possibly the greater environmental awareness linkedto the widespread discussion of the Protocol) have led to greater RETdiffusion In additional estimations (available upon request) the possibilityof time-specific effects was further investigated by including year and five-year period dummies Both variations showed that RE sector developmenthas picked up markedly since the mid-1990s and the trend even acceleratedafter 2000 lending further support to the idea that the Kyoto Protocol hasindeed had positive effects

Per capita income has the expected positive sign and is often highly signi-ficant confirming that richer countries also produce more energy The FDIshares of GDP however have no consistent sign being positive and highlysignificant in columns (1) and (4) and negative and insignificant otherwiseFinally there appear to be significant differences in Middle Eastern andNorth African countries (mena) which include most OPEC countries andsub-Saharan African countries (ssa) both regions have lower RE electricitygeneration on average than other non-OECD countries Non-hydro RE useis significantly lower especially in sub-Saharan Africa and Europe andCentral Asia (eca) in the latter case probably due to Soviet-era policies

The first results are encouraging but still leave room for some omittedvariables bias the explanatory power given by the R-squareds though notunusually low for such a large and varied sample of countries is modestIn table 3 we therefore consider additional variables We control for overallinstitutional quality (econfree) and the effects of oil production and pricesFor space reasons we concentrate on the two narrower financial sectorvariables dbacba (panel A) and pcrdbgdp (panel B) and do not show resultsfor hydropc (which are very similar to those for repc) as well as for the controlvariables income per capita FDIGDP and the regional dummies

Panel A shows that commercial banking (dbacba) continues to have astrong positive effect on RE use The coefficients in columns (1)ndash(3) for alltypes of RETs remain in the same order of magnitude as in table 2 howeverthe coefficients in columns (4)ndash(6) increase remarkably with beta coefficientsfor a one-standard-deviation change in commercial banking developmentof 029 This indicates that this particular area of the financial sector can havea real impact on the adoption of non-hydro RETs even once we take intoaccount the policy and institutional environment The same can howevernot be said for credit to the private sector (pcrdbgdp) in panel B which is notrobust to controlling for institutional quality (econfree) and oil productionand prices17

17 In additional estimations with the institutional quality measures compiled byKaufman et al (2008) pcrdbgdp proves significant at the five per cent level for repcand hydropc but not geopc Results available upon request

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 17: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 257

Table 3 Robustness analysis with additional regressors

(1) repc (2) repc (3) repc (4) geopc (5) geopc (6) geopc

Panel Adbacba 0086b 0086b 0087b 0025c 0025c 0025c

(234) (234) (237) (607) (609) (599)psreform minus0102b minus0101b minus0101b 0001 0001 0001

(200) (197) (198) (054) (073) (073)kyoto 0022b 0022b 0023b 0005c 0004c 0004c

(205) (204) (212) (374) (360) (331)econfree 0013a 0013a 0014a 0001 0001 0001

(181) (181) (184) (139) (122) (098)oilprodpc minus0081 minus0057 minus0043b minus0026

(021) (015) (219) (116)oilprice minus00004 00001a

(087) (181)oilprodpclowast minus0004 minus0001a

oilprice (079) (172)Observations 1829 1829 1829 1829 1829 1829Countries 92 92 92 92 92 92R2 within 004 004 004 010 010 010R2 between 009 009 009 005 007 006R2 overall 009 009 009 008 008 008

Panel Bpcrdbgdp 0053 00538 005 minus0004 minus0005 minus0005

(128) (129) (118) (080) (101) (103)psreform minus0105b minus0105b minus0105b 0001 0001 0001

(196) (196) (197) (085) (105) (107)kyoto 0027b 0027b 00294c 0005c 0005c 0004c

(249) (249) (268) (390) (379) (342)econfree 0020c 0020c 0022c 0004c 0003c 0003c

(283) (282) (299) (410) (408) (380)oilprodpc 0046 0058 minus0036a minus0019

(014) (017) (183) (086)oilprice minus0001 00001b

(149) (202)oilprodpclowast minus0001 minus0001

oilprice (024) (162)Observations 1756 1756 1756 1756 1756 1756Countries 89 89 89 89 89 89R2 within 003 003 004 007 007 007R2 between 011 011 011 013 016 016R2 overall 010 010 010 011 012 012

Notes All regressions are random-effects GLS on sample panel of non-OECDcountries with one-year-lags for all indicators except (psreform) kyoto and theregional dummies (not shown) Also not shown are the coefficients for gdppcand fdigdp as well as the constant term Absolute z-statistics in parentheses ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 18: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

258 Christa N Brunnschweiler

The measure of economic freedom (econfree) has the expected positivesign and is mostly significant especially in panel B indicating that astable institutional framework positively affects investments in the REpower sector Oil production and oil prices seem to have most impacton the use of non-hydro RETs (columns (4)ndash(6)) oil producers have lesselectricity generated with wind solar geothermal and other non-hydro REAn oil price increase on the other hand makes investment in these RETsmore worthwhile indicated by the positive sign However the interactionterm oilprodpc times oilprice shows that this lsquosubstitution effectrsquo away fromconventional fossil fuels to RETs is less pronounced in oil-producingcountries and may even be completely cancelled out For example at thesample mean oil production of 004 barrels per day and per capita anoil price increase of one standard deviation (1131 US$ per barrel) would ndashother things equal ndash lead to an increase in the non-hydro RE use in electricitygeneration of 0034 standard deviations a negligible effect18 However thesame price increase in the relatively largest oil producer (Qatar with 211barrels per day and per capita) would hypothetically lead to a massivedecrease in the use of non-hydro RE of over one standard deviation

Table A of the appendix shows results using coal and natural gas Thepattern for natural gas (panel B) is similar to that for oil while the effectsof coal price increases are consistent across all countries coal producersand otherwise (panel A) This suggests that (major) oil and natural gasproducers generally have lower levels of RE use probably reflectingdifferent investment incentives

In additional estimations shown in table B in the appendix we foundthat the economic transition of former East Bloc countries had a weaklysignificant impact on RE generation power production with all typesof RE has increased steadily since the transition shock The financialmarkets shock of 1997ndash1998 however had no strong impact on overall REproduction in developing and transition countries but non-hydro RET usehas increased significantly since the financial crisis of the late 1990s Notehowever that this positive impact may be combined with a post-Kyotoeffect since the time periods coincide in fact the simple Kyoto dummyvariable ndash which is otherwise consistently highly significant ndash loses itsstrength here Our measures of RE potential prove inconclusive both havepositive signs but neither approaches conventional levels of significance

Finally we compare different samples in table C starting with theentire world in panel A Credit to the private sector (pcrdbgdp) is the onlysignificantly positive financial sector variable while financial depth (llgdp)once again shows a significant negative impact on non-hydro RE use Notethat the effect of (pcrdbgdp) is much higher than we saw previously fornon-OECD countries and strongly significant for all types of RE while atthe same time financial depth (llgdp) shows an increased negative effect ongeopc

Interestingly there appears to be no consistent post-Kyoto effect inthe entire world sample this may point to policies on RE which were

18 Using results in panel A ((004 times 1131 times (minus0001)) + (1131 times 00001)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 19: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 259

already being enforced prior to the adoption of the Kyoto ProtocolFor example according to the World Energy Councilrsquos Survey of EnergyResources of 1998 many OECD countries were already close to realisingtheir technically exploitable hydropower capacity at the time They wereprobably farther away from their capacity frontiers as regards non-hydroRETs the Kyoto dummy is marginally significant in column (9) indicatingthe possibility of positive effects of the commitments undertaken within theKyoto Protocol However the effect is too weak to draw any more definiteconclusions

Panel B includes only low- to upper-middle-income non-OECD countries(no high-income) the main effects from tables 2ndash3 are confirmed and theexplanatory power increases substantially Panel C shows results for non-OECD countries without outliers Paraguay (as regards overall RE and hydroelectricity generation) and Costa Rica (for non-hydro production) whichagain confirm the main results Overall the positive impact of financialsector development on RE and especially non-hydro electricity generationseems remarkably robust Note that the negative effect of llgdp on geopcappears to be driven by the outlier Costa Rica as it disappears when wedrop that country

32 Dynamic panel data estimationsThe results of the panel estimations shown above are very suggestive ofa robust impact of financial sector development ndash especially commercialbank asset share and private credit share ndash on RE production in non-OECDcountries However the explanatory power of the specifications is generallybetween 9 and 20 per cent of the overall sample variation indicating somepossible omitted variable bias moreover energy sector development maydisplay a dynamic development over time In order to address these pointsand also confirm the causal relationship we perform dynamic one-stepGMM estimations after Arellano and Bond (1991)19 Table 4 presents theresults of the basic regressions using equation (2) The highly significantlagged dependent variables suggest that a dynamic approach is justifiedBeyond this we note that the results from the dynamic estimations confirmthe positive and significant effect of financial sector development on REproduction A post-Kyoto (policy) effect is noticeable only for non-hydroRETs shown in columns (7)ndash(9) The magnitudes of the single impactsremain similar to those seen above with GLS though they are slightlyhigher for repc and hydropc and slightly lower for geopc

In table 5 we show results with additional variables econfree and oilproduction and prices similar to table 3 For space reasons we concentrateon commercial banking (dbacba) which again proves the most robustfinancial sector variable particularly as regards geopc Note that thecoefficients for the commercial banking share now approach the magnitudesseen in table 2 above However the oil variables no longer appear to haveany impact and in some cases even change signs There is also no longer

19 Tests performed after two-step GMM estimations with the same specificationsshow that the overidentifying restrictions are valid and that there is no indicationof second-order autocorrelation

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 20: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

260C

hristaN

Brunnschw

eiler

Table 4 Dynamic panel estimations

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0096c 0093c 0011c

(316) (307) (741)llgdp 0093a 0092a 0001

(195) (194) (055)pcrdbgdp 0116b 0122c minus000004

(245) (258) (minus0024)gdppc minus00000001 0000004 0000004 minus0000001 0000003 0000003 0000001c 0000001c 0000001c

(minus0043) (150) (146) (minus032) (127) (118) (361) (462) (482)fdigdp 0001b minus00001 minus00002 0001b minus00001 minus00001 minus000001 00000001 0000001

(227) (minus021) (minus025) (232) (minus019) (minus023) (minus061) (00034) (0040)kyoto 0008 0008 0006 0005 0006 0003 0001b 0001c 0001c

(099) (099) (073) (061) (069) (036) (256) (316) (338)Lrepc 0697c 0704c 0702c

(490) (482) (479)Lhydropc 0696c 0702c 0700c

(489) (480) (477)Lgeopc 0838c 0848c 0848c

(867) (849) (846)Observations 2703 2441 2427 2703 2441 2427 2703 2441 2427Countries 136 124 124 136 124 124 136 124 124

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses Constant term included in all specifications (not shown) a b c statisticallysignificant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 21: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environm

entandD

evelopmentE

conomics

261

Table 5 Dynamic panel estimations with additional regressors

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

dbacba 0083a 0083a 0083a 0084a 0085a 0085a 0015c 0015c 0016c

(184) (185) (182) (189) (190) (188) (728) (728) (735)gdppc 0000002 0000002 0000002 0000002 0000002 0000002 00000004c 00000005b 00000004b

(064) (066) (059) (065) (067) (061) (261) (252) (214)fdigdp minus0001 minus0001 minus0001 minus0001 minus0001 minus0001 minus00001 minus00001 minus00001

(minus072) (minus075) (minus071) (minus063) (minus065) (minus062) (minus094) (minus091) (minus093)kyoto 0003 0003 0003 0001 0001 0001 0001 0001 0001

(031) (029) (026) (011) (0090) (0058) (135) (137) (138)econfree 0009 0008 0009 0003 0003 0005 00003 00003 00003

(106) (107) (123) (046) (046) (064) (092) (094) (081)oilprodpc minus0101 minus0173 minus0091 minus0162 0013 0017

(minus026) (minus044) (minus024) (minus042) (052) (067)oilprice minus00003 minus00003 000002

(minus056) (minus057) (103)oilprodpclowast 0008 0007 minus00003

oilprice (090) (082) (minus062)Lrepc 0698c 0698c 0698c

(426) (426) (425)Lhydropc 0696c 0696c 0696c

(425) (425) (424)Lgeopc 0827c 0827c 0826c

(710) (709) (702)Observations 1907 1907 1907 1907 1907 1907 1907 1907 1907Countries 100 100 100 100 100 100 100 100 100

Notes All regressions are ArellanondashBond dynamic panel regressions using sample panel of non-OECD countries with one-year-lagsfor all indicators Absolute z-statistics in parentheses gdppc fdigdp and constant term included in all specifications (not shown) a b c

statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 22: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

262 Christa N Brunnschweiler

any observable lsquoKyoto effectrsquo on RET use in electricity production Overallthe results for the dynamic GMM estimations prove weaker than forGLS although financial sector development ndash in particular the commercialbanking share ndash still has significant positive impacts on RE use

In sum the results of the empirical analysis support the idea thatfinancial intermediary development encourages the growth of the REsector However the effect is quite small in magnitude when we considerthe beta coefficients only for the case of non-hydro RE is the impacteconomically important throughout the specifications Moreover not allfinancial measures are equally important commercial banking seems todeliver the best support for the realisation of RE projects As regards policywe find an ambiguous impact of the power sector reform index it is difficultto explain this as unfortunately it is a time-invariant index and does notspecifically examine RE policies but looks at how level the playing fieldis for all types of resources and producers The more RE-relevant Kyotodummy however shows that there has been a clear positive developmentin RET use in transition and developing countries since the adoption of theKyoto Protocol in late 1997 This may be due to the growing number ofCDM projects or to a growing environmental awareness with domestic REpolicies or both Finally the findings are robust to the inclusion of othercovariates which could influence RE sector development and to variousdifferent sample sizes

4 Discussion and conclusionsThis paper examines the effects of financial intermediation on thedevelopment of the RE sector in a series of panel data estimations forthe period 1980ndash2006 Energy production today relies on exhaustible andpolluting conventional fossil fuels and a larger share of alternative energysources in primary energy production would not only have positiveenvironmental effects but would also bring greater energy security forfuture generations as RETs exploit domestic RE resources The increaseduse of RETs is one of the instruments to achieve the emission-reductiongoals of the Kyoto Protocol and it is also mentioned for example by the UNas a crucial part of achieving sustainable development

The focus is on non-OECD developing and transition countries Energyfirms in less developed economies are largely dependent on externalfinancing to realise new projects in turn external financing in thesecountries relies on the banking sector as stock and bond markets as well asventure capitalism are not well enough established to provide large-scalefunding However the underdevelopment of the banking sector in additionto specific RE-sector problems such as high up-front and information costsand long lead times hamper the emergence of RE entrepreneurs Thefinancing problems are combined with the greater issue of energy sectorregulations and RET policies which do not always offer a level playingfield for all energy producers as fossil fuel generation often benefits fromspecial incentives as well as the advantage of well-established technologiesand hence fewer unknowns for potential investors

The empirical estimations using RE electricity generation per capita as aproxy for RE sector development show that financial sector development

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 23: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 263

does indeed have a robust and significant positive effect on the amount ofRE produced which is independent of (or in addition to) energy policyWe found that the effect is particularly large when we consider non-hydropower RE (ie geothermal solar wind wood and waste) wherea one-standard-deviation increase in our financial intermediation measuresleads to an increase in non-hydro RE of up to 03 of a standard deviation Ofthe three financial sector measures used the commercial bank asset shareand the private credit share proved the most robust while financial depth isprobably too broad a measure to adequately capture the more bank-focuseddevelopment that is assumed to be important in developing and emergingeconomies The results are robust to controlling for additional effects andto different sample specifications

The findings also suggest that there has been a strong positive post-Kyotoeffect in RET use in non-OECD countries though we cannot pinpoint theexact cause to either the Kyoto Protocol mechanisms such as CDM and JI ora growing environmental awareness with consequent domestic RE policiesThe power sector reforms indicator had an ambiguous impact on RE sectordevelopment the effect on overall RE use and hydropower tended to benegative while the influence on non-hydro RETs was more positive Thelack of a clearer conclusion in this regard may be due to the data the powersector reform index provided by ESMAP (1999) is available for only oneyear (1998) and concentrates on evaluating competitiveness and creatinga level playing field for all types of energy resources with no particularfocus on RETs Moreover the mixed reform results to date in non-OECDcountries have led to a general rethinking of the objectives and underlyingassumptions of power sector reform (eg Williams and Ghanadan 2006)which will hopefully lead to more adequate measures (and incentives) ofreform progress

As regards possible differences in fossil fuel producing countries there isindeed some indication that (major) oil and natural gas producers generallysee less electricity generation with RETs particularly non-hydro RE thanother countries Fossil fuel price increases likely weaken the incentives toinvest in RETs in major producers while their effect is opposite in smallerproducers and fossil fuel-poor countries

The approach offers a first attempt at empirically verifying the role offinance for the development of the RE industry The availability of qualitydata on RE development and investment has so far hampered empiricalstudies in this area further work is needed to corroborate the resultsespecially in the form of case studies Additionally it will be interestingto see the future impacts and developments of the Kyoto Protocol

Any policy recommendations must remain tentative at this pointNevertheless it seems safe to say that the financial sector does indeedhave a measurable impact on the emergence of RE producers Aregulatory framework aimed at fostering the RE sector cannot neglect thefinancing aspects and particularly the availability of private sector financialintermediation Developing and strengthening the financial sector of coursehas greater macroeconomic benefits as demonstrated by the vast finance-and-growth literature however it also has a non-negligible influence onthe success of RE especially the most recent non-hydro RETs

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 24: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

264 Christa N Brunnschweiler

ReferencesArcher CL and MZ Jacobson (2005) lsquoEvaluation of global wind powerrsquo Journal

of Geophysical Research 110 D12110Arellano M and S Bond (1991) lsquoSome tests of specification for panel data Monte

Carlo evidence and an application to employment equationsrsquo Review of EconomicStudies 58(2) 277ndash97

Babbar S and J Schuster (1998) lsquoPower project finance experience in developingcountriesrsquo World Bank RMC Discussion Paper No 119

Baltagi BH (2008) Econometric Analysis of Panel Data 2nd Edn Chichester UK JohnWiley amp Sons

Barbier E (2009) A Global Green New Deal Report prepared for the Economics andTrade Branch Division of Technology Industry and Economics UNEP February2009

Beck T A Demirguc-Kunt and R Levine (2000) lsquoA new database on financialdevelopment and structurersquo World Bank Economic Review 14 597ndash605

Beck T A Demirguc-Kunt L Laeven and V Maksimovic (2004a) lsquoThedeterminants of financing obstaclesrsquo Journal of International Money and Finance25(6) 932ndash952

Beck T A Demirguc-Kunt and V Maksimovic (2004b) lsquoBank competition andaccess to finance international evidencersquo Journal of Money Credit and Banking36(3) 627ndash648

Brunnschweiler CN (2006) lsquoFinancing the alternative renewable energy indeveloping and transition countriesrsquo Economics Working Paper Series 0649ETH Zurich Zurich Switzerland

Carlin W and C Mayer (2003) lsquoFinance investment and growthrsquo Journal of FinancialEconomics 69(1) 191ndash226

Churchill AA and RJ Saunders (1989) lsquoFinancing of the energy sector indeveloping countriesrsquo World Bank Industry and Energy Department WorkingPaper Energy Series Paper No 14

Deffeyes KS (2005) Beyond Oil The View from Hubbertrsquos Peak New York Hill andWang

Demirguc-Kunt A and R Levine (2001) Financial Structure and Economic GrowthA Cross-Country Comparison of Banks Markets and Development Cambridge MAMIT Press

Demirguc-Kunt A and V Maksimovic (1999) lsquoInstitutions financial markets andfirm debt maturityrsquo Journal of Financial Economics 54 295ndash336

Economist The (2009) lsquo2020 vision The peak oil debatersquo December 10thEIA (Energy Information Administration) (2000) lsquoLong term world oil

supply presentation prepared for the American Association of Petro-leum Geologistsrsquo [Online] httpwwweiadoegovpuboil_gaspetroleumpresentations2000long_term_supplyindexhtm (posted July 28 2000)

EIA (Energy Information Administration) (2006) lsquoInternational energy annualrsquo[Online] httpwwweiadoegov (release December 8 2008)

ESMAP (1999) lsquoGlobal energy sector reform in developing countries a scorecardrsquoVol 1 World Bank ESM219 (July 99)

Gwartney J R Lawson J Hall and S Norton (2008) Economic Freedom of theWorld 2008 Annual Report Economic Freedom Network

Head C (2000) lsquoFinancing of private hydropower projectsrsquo World Bank DiscussionPaper No 420

Huang Y and T Barker (2009) lsquoThe Clean Development Mechanism and sustainabledevelopment a panel data analysisrsquo Tyndall Centre for Climate Change ResearchWorking Paper no 130

IEA (International Energy Agency) (2003) World Energy Investment Outlook 2003Insights Paris International Energy Agency

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 25: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 265

IEA (International Energy Agency) (2008) World Energy Outlook 2008 ParisInternational Energy Agency

Ishiguro M and T Akiyama (1995) lsquoElectricity demand in Asia and the effectson energy supply and the investment environmentrsquo World Bank Policy ResearchWorking Paper No 1557

Kaufmann D A Kraay and M Mastruzzi (2008) lsquoGovernance matters VIIGovernance indicators for 1996ndash2007rsquo World Bank Policy Research Working PaperNo 4654

King RG and R Levine (1993)a) lsquoFinance and growth schumpeter might be rightrsquoQuarterly Journal of Economics 108 717ndash738

King RG and R Levine (1993)b) lsquoFinance entrepreneurship and growthrsquo Journalof Monetary Economics 32 513ndash542

Levine R and S Zervos (1998) lsquoStock markets banks and economic growthrsquo TheAmerican Economic Review 88(3) 537ndash558

Levine R N Loayza and T Beck (2000) lsquoFinancial intermediation and growthcausality and causesrsquo Journal of Monetary Economics 46(1) 31ndash77

MacLean JC and JM Siegel (2007) lsquoFinancing Mechanisms and PublicPrivateRisk Sharing Instruments for Financing Small Scale Renewable EnergyEquipment and Projectsrsquo report prepared for UNEP and GEF [Online]httpwwwuneptieorgenergyactivitiesfrmpdfSSREpdf

Pacudan R (2005) lsquoThe Clean Development Mechanism new instrument infinancing renewable energy technologiesrsquo in A Iacomelli (ed) Renewable Energiesfor Central Asia Countries Economic Environmental and Social Impacts Springerpp 27ndash42

Painuly JP and N Wohlgemuth (2006) lsquoRenewable energy financing - what canwe learn from experience in developing countriesrsquo Energy Studies Review 14(2)154ndash170

Renner M S Sweeney and J Kubit (2008) Green Jobs Towards a Decent Work in aSustainable Low-carbon World Geneva UNEPILOIOEITUC

Sonntag-OrsquoBrien V and E Usher (2004a) lsquoFinancing options for renewableenergyrsquo Environmental Finance Renewables Finance special issue [online] httpwwwenvironmental-financecom20040504mayfinanchtm (posted May 2004updated July 4 2008)

Sonntag-OrsquoBrien V and E Usher (2004b) lsquoMobilising Finance for RenewableEnergy Thematic background paper for the International Conference for Renew-able Energiesrsquo Bonn [Online] httpwwwrenewables2004depdftbdTBP05-financingpdf January 2004

Tadesse S (2002) lsquoFinancial architecture and economic performance internationalevidencersquo Journal of Financial Intermediation 11 429ndash454

Tharakan PJ J de Castro and T Kroeger (2007) lsquoEnergy sector assistance indeveloping countries current trends and policy recommendationsrsquo Energy Policy35 734ndash738

Tirpak D and H Adams (2008) lsquoBilateral and multilateral financial assistance forthe energy sector of developing countriesrsquo Climate Policy 8 135ndash151

UN Commission on Sustainable Development (2007) Report on the fifteenth sessionEconomic and Social Council Official Records Supplement No 9

UNEP FI (2004) Renewable Energy United Nations Environment ProgrammeFinance Initiative Climate Change Working Group CEO Briefing (June2004)

van Hulle F G van Roekel J de Jongh J Borchgrevink J C Lopez and E F Vindas(2003) Evaluation of the environmental impact of the 20MW Tejona wind powerproject Costa Rica ECN report no ECN-C-02-93

World Energy Council (WEC) (1980ndash2007) Survey of Energy Resources London WorldEnergy Council

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 26: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

266 Christa N Brunnschweiler

Williams JH and R Ghanadan (2006) lsquoElectricity reform in developing andtransition countries a reappraisalrsquo Energy 31 815ndash844

Wohlgemuth N and JP Painuly (1999) lsquoPromoting private sector financingof commercial investments in renewable energy technologies in Finance forSustainable Development Testing New Policy Approaches United Nations

World Bank (1999) Energy and Development Report Washington DC IBRDWorldBank

World Bank (2002) Financing for Sustainable Development Washington DCIBRDWorld Bank

World Bank (2003) Power for Development A Review of the World BankGrouprsquos Experience with Private Participation in the Electricity Sector OperationsEvaluation Department Washington DC IBRDWorld Bank

Appendix Definitions and sourcesAll data were collected for non-OECD countries (as of 1980 ndash the recentOECD members Czech Republic Hungary Korea Mexico Poland andSlovakia were included in the estimations) and for OECD countries forcomparison for the years 1980ndash2006 (where available)

Variable Definition Source

repc Net total renewable resource electricpower generation in billion kwhper capita ndash including hydrowood and waste geothermalsolar and wind

Own calculationfrom EIA (2006)

hydropc Net hydroelectric power generationin billion kwh per capita

Own calculationfrom EIA (2006)

geopc Net non-hydro renewable resourceelectric power generation inbillion kwh per capita ndash includinggeothermal wind solar andwood and waste

Own calculationfrom EIA (2006)

dbacba Deposit money bank assets (deposit money + central) bankassets (ie commercial bank assetshare versus Central Bank)

Beck et al (2000)

pcrdbgdp Private credit by deposit moneybanksGDP

Beck et al (2000)

llgdp Liquid LiabilitiesGDP (financialdepth)

Beck et al (2000)

psreform Qualitative power sector reformindicator for 1998 ranging fromzero (no reforms) to six (allrelevant reforms implemented inall areas) All OECD and otherhigh-income countries notincluded in the original studywere assigned a value of 6

ESMAP (1999)

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 27: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environment and Development Economics 267

Variable Definition Source

kyoto Dummy variable taking value onefrom 1998 onwards (post-KyotoProtocol period)

Own construction

gdppc GDP per capita (PPP adjusted) incurrent US$

World DevelopmentIndicators (WDI)

econfree Economic freedom summary index Gwartney et al (2008)fdigdp Foreign direct investment net

inflows (GDP)WDI

oilprodpc Per capita oil production in barrelsper day

Own calculation fromEIA (2006)

oilprice Crude oil prices measured in US$per barrel in current dollars

British Petroleum

natgasprodpc Per capita Natural gas production inbtu equivalents

British Petroleum

natgasprice Natural gas price in US$ per millionbtu in current dollars

British Petroleum

coalprodpc Per capita coal production in milliontonnes

British Petroleum

coalprice Coal price in US$ per tonne incurrent dollars

British Petroleum

transition Measure for transition shock inCentral and Eastern Europe aswell as former Soviet countriesYears until 1991 have value zeroand post-transition years take onincreasing values All othercountries receive value zerothroughout the period

Own construction

fincrisis Financial crisis measure takingvalue zero until 1997 thenincreasing with every post-crisisyear

Own construction

eca Dummy variable for countries ofEurope and Central Asia

mena Dummy variable for countries of theMiddle East and Northern Africa

ssa Dummy variable for countries ofsub-Saharan Africa

terms of use available at httpswwwcambridgeorgcoreterms httpsdoiorg101017S1355770X1000001XDownloaded from httpswwwcambridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cambridge Core

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 28: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

268C

hristaN

Brunnschw

eiler

Table A Robustness analysis with coal and natural gas production

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0102c 0086c 0015c

(340) (290) (474)pcrdbgdp 0120c 0110b 0008

(265) (246) (163)llgdp 0057 0058 minus0004

(124) (129) (083)psreform minus0098b minus0101b minus0101b minus0099b minus0102b minus0103b 0001 0001 0001

(227) (213) (214) (230) (217) (218) (073) (093) (100)kyoto 0028c 0033c 0034c 0025c 0028c 0029c 0004c 0005c 0005c

(351) (373) (380) (321) (328) (333) (443) (545) (573)coalprodpc 0003 minus0021 minus0015 0011 minus001 minus0005 minus0004 minus0003 minus0003

(01) (055) (039) (037) (026) (013) (160) (138) (115)coalprice 00002 00004 00004 00002 00004 00003 00001 00001b 00001b

(065) (117) (101) (055) (096) (080) (148) (233) (236)coalprodpclowast 00002 00004 00003 00002 00003 00003 00001 00001 00001

coalprice (078) (109) (106) (064) (092) (090) (160) (162) (153)Observations 1965 1756 1770 1965 1756 1770 1965 1756 1770Countries 119 107 107 119 107 107 119 107 107r2within 004 003 002 003 002 002 008 007 00600r2between 012 012 012 012 012 012 008 010 011r2overall 010 011 011 010 011 011 009 009 010

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 29: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environm

entandD

evelopmentE

conomics

269

Panel Bdbacba 0110c 0093c 0016c

(410) (352) (537)pcrdbgdp 0093b 0088b 0003

(235) (227) (079)llgdp 0043 0050 minus0009b

(104) (124) (198)psreform minus0097b minus0101b minus0101b minus0097b minus0102b minus0102b 0001 0001 0001

(226) (216) (215) (227) (218) (217) (075) (111) (128)kyoto 0023c 00304c 0032c 0021c 0026c 0027c 0003c 0005c 0005c

(304) (356) (370) (277) (314) (325) (387) (480) (512)natgasprodpc minus0152 minus0115 minus0179 minus0147 minus0099 minus0163 minus0035 minus0036 minus0035

(044) (034) (052) (044) (029) (049) (101) (100) (098)natgasprice 0003 0004 0003 0002 0003 0002 0001b 0001c 0001c

(084) (108) (093) (067) (077) (059) (199) (294) (316)natgasprodpclowast minus0002 0002 0009 0005 0006 0014 minus0001 minus0001 minus0003

natgasprice (004) (003) (015) (009) (011) (025) (025) (017) (040)Observations 2178 1943 1956 2178 1943 1956 2178 1943 1956Countries 119 107 107 119 107 107 119 107 107R2 within 004 003 002 003 002 002 008 006 006R2 between 012 012 012 012 012 012 008 013 014R2 overall 010 011 011 010 011 011 010 011 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Absolute z-statistics in parenthesesa b c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 30: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

270C

hristaN

Brunnschw

eiler

Table B RE potential and time shock variables

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel Adbacba 0101c 0080c 0016c

(343) (279) (582)pcrdbgdp 0055 0062 0002

(133) (150) (059)llgdp 0029 005 minus0009b

(063) (111) (210)psreform minus0099b minus0101a minus0100a minus0099b minus0102a minus0101a 0001 0001 0001

(197) (183) (183) (200) (188) (188) (062) (103) (118)kyoto 0021c 0033c 0032c 0019b 0029c 0028c 0004c 0006c 0006c

(258) (352) (348) (234) (315) (305) (520) (651) (686)hydropot 00001 000005 000005 00001 000005 000005

(132) (053) (054) (133) (057) (057)windpot 0049 0039 0038 0006 0006 0006

(018) (014) (014) (118) (111) (115)Observations 2157 1900 1913 2157 1900 1913 2450 2179 2192Countries 101 93 93 101 93 93 119 107 107R2 within 005 004 0042 004 005 003 007 005 005R2 between 012 013 013 012 012 012 009 014 014R2 overall 011 012 012 010 011 011 010 011 011

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 31: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environm

entandD

evelopmentE

conomics

271

Panel Bdbacba 0084c 0068c 0015c

(337) (275) (543)pcrdbgdp 0086b 0082b 0003

(243) (238) (080)llgdp 0048 0058 minus001b

(132) (162) (245)psreform minus0087b minus0087a minus0087a minus0087b minus0089a minus0089a 0001 0002 0002a

(201) (184) (185) (203) (minus188) (minus188) (110) (155) (166)kyoto 002a 0026b 0027b 0019a 0024b 0026b 0001 0002 0002

(184) (218) (233) (178) (208) (223) (098) (133) (134)fincrisis 0001 0002 0002 0001 0002 0001 0001c 0001c 0001c

(062) (114) (093) (038) (078) (053) (271) (340) (379)transition 0003 0005a 0005a 0002 0004 0004 0001b 0001c 0001c

(116) (165) (188) (094) (126) (155) (242) (363) (315)Observations 2439 2169 2182 2439 2169 2182 2439 2169 2182Countries 118 106 106 118 106 106 118 106 106R2 within 004 003 003 003 002 002 008 007 007R2 between 010 010 010 010 010 010 010 015 016R2 overall 009 009 009 008 009 009 011 012 012

Notes All regressions are random-effects panel regressions using sample panel of non-OECD countries with one-year-lags for allindicators with the exception of psreform kyoto fincrisis and transition Constant term psreform kyoto gdppc fdigdp and regionaldummies not shown for space reasons Absolute z-statistics in parentheses Constant term included in all specifications (not shown) ab c statistically significant at 10 5 and 1 per cent levels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 32: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

272C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel A worlddbacba 0047 0069 minus0035

(050) (100) (105)pcrdbgdp 0819c 0417c 0368c

(898) (616) (117)llgdp minus0087 0015 minus0141c

(076) (018) (378)psreform 0100 0114 0130 0119 0134 0142 minus0011 minus0015 minus0006

(075) (078) (091) (092) (096) (103) (116) (144) (053)kyoto 0014 0015 0035 0018 002 0029 0009 0007 0018a

(055) (054) (118) (093) (092) (136) (094) (071) (170)Observations 3014 2752 2757 3014 2752 2757 3014 2752 2757Countries 142 130 130 142 130 130 142 130 130R2 within 007 009 006 003 004 003 013 018 013R2 between 009 009 009 007 007 007 012 011 017R2 overall 012 010 012 009 008 008 015 012 017

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 33: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

Environm

entandD

evelopmentE

conomics

273

Panel B no hi-incomedbacba 0063b 005a 0012c

(233) (188) (380)pcrdbgdp 0084b 0084b 0001

(199) (202) (012)llgdp minus0002 0017 minus0016c

(004) (040) (335)psreform minus0092a minus0095a minus0095a minus0094a minus0097a minus0097a 0001 0002a 0002a

(183) (167) (166) (185) (170) (169) (137) (180) (180)kyoto 0006 0015 0016a 0005 0013 0013 0002a 0003c 0003c

(076) (156) (166) (067) (136) (142) (177) (259) (294)Observations 2179 1879 1892 2179 1879 1892 2179 1879 1892Countries 103 91 91 103 91 91 103 91 91R2 within 006 006 006 005 004 004 010 009 010R2 between 015 015 015 014 014 014 022 026 025R2 overall 013 014 014 013 013 013 018 020 019

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core

Page 34: Finance for renewable energy: an empirical analysis of developing …doc.rero.ch/record/303054/files/S1355770X1000001X.pdf · 2017-08-02 · 242 Christa N. Brunnschweiler and possibly

274C

hristaN

Brunnschw

eiler

Table C Outlier analysis and world comparison

(1) repc (2) repc (3) repc (4) hydropc (5) hydropc (6) hydropc (7) geopc (8) geopc (9) geopc

Panel C no outliersdbacba 0095c 0077c 0007c

(540) (457) (443)pcrdbgdp 0049b 0046b 0002

(202) (201) (090)llgdp 0048a 0057b 0001

(196) (241) (062)psreform minus0025 minus0020 minus0020 minus0025 minus0021 minus0021 0001 0001 0001

(110) (086) (087) (112) (091) (093) (093) (130) (127)kyoto 0014c 0025c 0024c 0011b 0019c 0018c 0002c 0003c 0003c

(276) (464) (449) (217) (376) (355) (475) (619) (603)Observations 2433 2162 2175 2433 2162 2175 2424 2153 2166Countries 118 106 106 118 106 106 118 106 106R2 within 006 004 0040 005 003 003 009 007 007R2 between 022 024 024 023 025 024 010 014 014R2 overall 017 019 019 017 019 018 012 012 012

Notes All regressions are random-effects panel regressions fdigdp gdppc regional dummy variables and constant term included butnot shown One-year-lags for all indicators with the exception of psreform and regional dummies Panel A shows world sample Panel Bshows sample of non-OECD without high-income countries Panel C shows sample of non-OECD countries without outliers Paraguay(columns 1ndash6) and Costa Rica (columns 7ndash9) Absolute z-statistics in parentheses a b c statistically significant at 10 5 and 1 per centlevels respectively

terms of use available at httpsw

ww

cambridgeorgcoreterm

s httpsdoiorg101017S1355770X1000001XD

ownloaded from

httpsww

wcam

bridgeorgcore University of Basel Library on 11 Jul 2017 at 155945 subject to the Cam

bridge Core