Policy Applications of Environmental Accounting

E N V I R O N M E N TA L E C O N O M I C S S E R I E S

Policy Applicationsof EnvironmentalAccounting

PAPER NO. 88

Glenn-Marie Lange

January 2003

Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The useand citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed tothe World Bank. Copies are available from the Environment Department, The World Bank, Room MC-5-126.

Policy Applicationsof EnvironmentalAccounting

Glenn-Marie Lange

January 2003

The financial assistance of the Governmentof Norway is gratefully acknowledged.

THE WORLD BANK ENVIRONMENT DEPARTMENT

The International Bank for Reconstructionand Development/THE WORLD BANK1818 H Street, N.W.Washington, D.C. 20433, U.S.A.

Manufactured in the United States of AmericaFirst printing January 2003

Glenn-Marie Lange is Research Associate Professor at the Institute for Economic Analysis, NewYork University.

iiiEnvironmental Economics Series

Contents

Chapter 1Introduction 1

Chapter 2Methodological Approaches to Environmental Accounting 3

2.1 Concepts of sustainability 32.2 Asset accounts 4

Depletion and depreciation 52.3 Pollution and material flow accounts 6

Physical accounts 6Monetary accounts for environmental degradation 6

2.4 Environmental protection and resource management accounts 72.5 Macroeconomic indicators 8

Physical indicators 8Monetary indicators 10

Chapter 3Asset Accounts 13

3.1 Monitoring total wealth and changes in natural capital 133.2 Managing resources—economic efficiency, sustainability, and other socio-economic

objectives 203.2.A Economic efficiency 213.2.B Sustainability 243.2.C Other socio-economic objectives 25

Chapter 4Physical Flow Accounts for Pollution and Material Use 27

4.1 Physical flow accounts 274.1.A Indicators and descriptive statistics 274.1.B Policy analysis and strategic planning 32

4.2 Monetary flow accounts for environmental degradation and resource use 364.2.A Indicators and descriptive statistics 364.2.B Policy analysis with monetary accounts 40

Chapter 5Environmental Protection and Resource Management Accounts 43

5.1 Indicators and descriptive statistics 43

iv Environment Department Papers

Biological Resource Management — Integrating Biodiversity Concerns in Rural Development Projects and Programs

Environmental protection expenditure accounts 43Resource management expenditures 45Environmental protection industry 45Environmental and resource taxes 45

5.2 Policy analysis 47

Chapter 6Economy-Wide Indicators of Sustainable Development 49

6.1 Physical indicators of macro-level performance 496.2 Environmentally-adjusted NDP and related indicators 50

Genuine savings 526.3 Modeling approaches to macro-economic indicators 52

Chapter 7Concluding Comments 57

Underutilization of accounts 57Comprehensive environmental accounts 57

International comparability 58Full assessment of environmental impact 58Progress toward international comparability 58

NOTES 59

REFERENCES 59

FIGURES

Figure 1. Biomass of hake, pilchard, horse mackerel in Namibia, 1963 to 1999 14Figure 2. Value of produced and non-produced assets in Australia, Botswana and Namibia

in current prices 16Figure 3. Index of per capita national wealth in Australia, Botswana, and Namibia,

1990 to 1998 19Figure 5. Resource rent and taxes from forestry in Norway, 1985 to 1995 22Figure 4. Resource rent and taxes from oil and gas mining in Norway, 1985 to 1996 22Figure 6. Resource rent and subsidies to fisheries in Norway, 1985 to 1995 23Figure 7. Direct and total emissions of sulfur dioxide per unit of industrial output delivered

to final users in Sweden, 1991 31Figure 8. Economic contribution and environmental burden from domestic pollution by

selected industries in Sweden, 1991 37Figure 9. Domestic emissions, exports and imports of NOX in Sweden, 1991 39Figure 10. Total environmental taxes and total direct subsidies, by industries and

final demand, 1995 46Figure 11. Index of macro-indicators for economic and environmental performance,

Netherlands, 1987 to 1998 50Figure 12. Percentage change in material use in five industrialized countries, 1975 to 1996 51

v

Contents

Environmental Economics Series

TABLES

Table 1. Countries with environmental accounting programs 2Table 2. Environmental macroeconomic indicators, physical and monetary 9Table 3. Natural capital as percent of total non-financial assets in Namibia, Botswana,

and Australia, 1990 to 1998 17Table 4. Financial, non-financial assets, and net worth in Botswana and Australia,

1990 to 1999 18Table 5. Cost of depletion of natural capital in the Philippines, 1988 to 1993 20Table 6. Index of water use, GDP growth and population growth in Botswana, 1993 to 1998 28Table 7. Net contribution of consumption and production to GDP and to six environmental

themes in the Netherlands, 1993 29Table 8. National income per cubic meter of water by sector in Botswana, Namibia, and

South Africa, 1996 30Table 9. Decomposition of the percent change in CO2 between 1987 and 1998,

the Netherlands 32Table 10. Emissions embodied in Swedish imports under alternative assumptions about

emission intensities of imports, 1995 36Table 11. Emissions of BOD and environmental damage by selected industries in the

Philippines, 1993 38Table 12. Cost of NOx emissions using different valuation methods in Sweden, 1991 39Table 13. Pollution damage by source country and receptor country in the European Union 40Table 14. Summary indicators of environmental protection expenditures in 1992 44Table 15. Environmental taxes in Sweden, 1993–98 46Table 16. eaNDP as percentage of NDP in selected countries 52Table 17. Genuine saving as percent of GDP, 1997 53Table 18. Hueting’s Sustainable National Income

1Environmental Economics Series

1 Introduction

The first environmental accounts (EA) wereconstructed by Norway in the 1970s and wereonly slowly adopted by other countries. In theearly 1990’s, the World Bank conducted areview of (EA), providing a compendium ofwhich countries had compiled environmentalaccounts, the methods that had been used toconstruct EA, and the extent of coverage (Peskinand Lutz, 1990). Since that time, EA haveincreasingly been recognized as a usefuleconomic tool, resulting in a great deal ofactivity in both developed and developingcountries. Over the last decade, conceptual andtechnical aspects of construction EA havereceived a great deal of attention; however,much less is known about how EA are beingused for policy.

The motivation for EA has been the adoption bygovernments, at least in principle, of the notionof sustainable development, coupled with theunderstanding that economic activities andappropriate economic incentives play a centralrole in determining whether development issustainable or not. EA provide policy-makers a)with indicators and descriptive statistics tomonitor of the interaction between theenvironment and the economy, and progresstoward meeting environment goals; and b) witha database for strategic planning and policyanalysis to identify more sustainabledevelopment paths, and the appropriate policyinstruments for achieving these paths.

This report reviews the policy applications ofEA in industrialized and developing countries,

and also indicates potential applications, whichmay not be fully exploited at this time. Thereport is intended to serve as a guide forcountries implementing EA by showing how EAcan support policy decision-making. It mayalso assist EA practitioners and scholars byproviding them with a better understanding ofthe needs of end-users of the accounts.

Table 1 identifies the major countries that areconstructing EA on an on-going basis in theirstatistical offices or other governmentministries. These countries have the mostextensive experience with policy use of the EAand provide the core of this report. Most of thework is being done in Europe, Australia, andCanada and a relatively few developingcountries. Of the developing countries, thePhilippines, Botswana, and Namibia areparticularly important because policy analysiswas built into the EA project design. There arecountless other one-time or academic studies; afew which are referred to in this report and arealso listed in Table 1.

The second section of the report provides a briefreview of the different approaches to EA,beginning with a discussion of the concepts ofsustainability that underlie different EAmethodologies. The methodologies arereviewed mainly as they relate to the System ofEnvironmental and Economic Accounts (SEEA),the handbook for EA developed by the UnitedNations, Eurostat, OECD, World Bank, andother agencies (UN, 1993, currently underrevision). The applications themselves are

Environment Department Papers2

Policy Applications of Environmental Accounting

described in four subsequent chaptersorganized around the structure of the EA: assetaccounts (section 3), flow accounts for materialsand pollutants (section 4), environmentalprotection and resource management

expenditures (section 5), and macroeconomicindicators (section 6). The final sectionprovides concluding remarks about the use ofEA for policy.

Table 1. Countries with environmental accounting programs

Notes: a accounts for water only.

Other European countries have also constructed environmental accounts but are not included here because of the limited

policy analysis of the accounts.

Flow accounts forpollutants & materials

Environmentalprotection &

resourcemanagement Macro

Assets Physical Monetary expenditures aggregates

Industrialized countries

Australia X X X

Canada X X X

Denmark X X X

Finland X X X

France X X X

Germany X X X X X

Italy X X X

Japan X X X X X

Norway X X

Sweden X X X X X

UK X X X

US X X

Developing countries

Botswana X X Xa

Chile X Xa X

Korea X X X X X

Mexico X X X X X

Moldova Xa

Namibia X X Xa

Philippines X X X X X

Occasional studies

Colombia X X X

Costa Rica X

Eu-15 X

Indonesia X

South Africa X X Xa


2Environmental and resource accounting evolvedsince the 1970s through the efforts of individualcountries or practitioners, each developing theirown frameworks and methodologies torepresent their environmental priorities. Sincethe early 1990s, concerted efforts have beenunderway through the United Nations StatisticsDivision, the European Union, the OECD, theWorld Bank, country statistical offices, andother organizations to standardize theframework and methodologies. The UnitedNations published an interim handbook onenvironmental accounting in 1993 (UN 1993),which is currently under revision. Thediscussion below describes the differentmethodologies and how they are related to therevised SEEA.

Environmental accounts have four components:• Natural resource asset accounts, which deal

mainly with stocks of natural resources andfocus on revising the Balance Sheets of theSystem of National Accounts (SNA)

• Pollutant and material (energy andresources) flow accounts, which provideinformation at the industry level about theuse of energy and materials as inputs toproduction and final demand, and thegeneration of pollutants and solid waste.These accounts are linked to the Supply andUse Tables of the SNA, which are used toconstruct input-output (IO) tables.

• Environmental protection and resourcemanagement expenditures, which identifiesexpenditures in the conventional SNAincurred by industry, government and

households to protect the environment ormanage resources

• Environmentally-adjusted macroeconomicaggregates, which include indicators ofsustainability such as environmentally-adjusted Net Domestic Product (eaNDP).

This section begins with a discussion ofconcepts of sustainability and the implicationsfor approaches to measuring sustainability, thendiscusses each component of the environmentalaccounts.

2.1 Concepts of sustainability

While this report cannot review all the literatureabout sustainability (see Pezzey (1989, 1994) forsuch a review), a brief discussion of the topic isnecessary in order to understand some of theissues underlying the different approaches toenvironmental accounting. The BrundtlandCommission Report, Our Common Future,popularized the notion of sustainabledevelopment as “…development that meets theneeds of the present without compromising theability of future generations to meet their ownneeds (WCED 1987). This rather vague conceptresonates with the economist’s basic notion ofsustainability, whose starting point has been theidea of income expressed by John Hicks“…income is the maximum amount anindividual can consume during a period andremain as well off at the end of the period as atthe beginning.” (Hicks 1946). Hicks’ statementhas generally been interpreted as the amount ofincome that can be spent without depleting thewealth which generates the income.

Methodological Approachesto Environmental Accounting



Hence, sustainability requires non-decreasinglevels of capital stock over time, or, at the levelof the individual, non-decreasing per capitacapital stock. Indicators of sustainability couldbe based on either the value of total assets everyperiod, or by the change in wealth, consumptionof capital (depreciation) in the conventionalnational accounts. For a proper measure ofsustainability, all assets should be included insuch an indicator: manufactured capital, naturalcapital and human capital. In the past, onlymanufactured capital was recorded in the SNA,but the recognition of the importance of naturalcapital has led to the expansion of the assetboundary to include this asset. (Human capitalhas not yet been included because there is noagreement about how to measure it and is notdiscussed further.)

Economic sustainability can be defined asstrong or weak, reflecting controversy over thedegree to which one form of capital cansubstitute for another. Weak sustainabilityrequires only that the combined value of allassets remain constant, that is, it is possible tosubstitute one form of capital for another, sonatural capital can be depleted or theenvironment degraded as long as there arecompensating investments in other types ofcapital: manufactured, human, or other type ofnatural capital.

Strong sustainability is based on the conceptthat natural capital is a complement tomanufactured capital, rather than a substitute.Renewable resources such as fish or forests, canbe exploited only at the natural rate of netgrowth; the use of non-renewable resourcesshould be minimized and, ideally, used only atthe rate for which renewable substitutes areavailable; emissions of wastes should notexceed the assimilative capacity of theenvironment. The indicator of sustainabilityrequires that all natural capital is measured inphysical units. A less extreme version of strongsustainability accepts some degree of

substitutability among assets, but recognizesthat there are some “critical” assets which areirreplaceable. The corresponding measure ofsustainability would be partly monetary (forthose assets, manufactured and natural, whichare not critical and for which substitution isallowed) and partly physical, for critical naturalassets.

Das Gupta and Maler (2000) have argued thatprices can fully reflect sustainability and thelimits to substitution. Hamilton (2000) pointsout the highly restrictive and unlikelyconditions that must be fulfilled in order forprices to provide a true measure ofsustainability.

2.2 Asset accounts

Natural resource asset accounts follow thestructure of the asset accounts of the SNA, withdata for opening stocks, closing stocks, andchanges during the year. The changes that occurduring the period are divided into those that aredue to economic activity (e.g., extraction ofminerals or harvesting of forests), and those thatare due to natural processes (e.g., growth,births, and deaths) or other factors. There issome controversy over how to treat newdiscoveries of minerals: as an economic change(the result of exploration activities), or as part ofother volume changes. The monetary accountsfor resources have an addition component, likemanufactured capital, for revaluation.

Measurement of the physical stocks can presentproblems both in terms of what to measure aswell as how to measure. In some earlier versionsof subsoil (mineral) asset accounts, onlyeconomically proven stocks were included inthe asset accounts. Some countries havemodified this to include a portion of probableand possible stocks, based on the probability ofthese stocks becoming economically feasible tomine. Certain resources, like marine capturefisheries are not observed directly and require


Methodological Approaches to Environmental Accounting

biological models to estimate stocks andchanges in stocks.

Two methods have been used to value assets:net present value (NPV) and net price. The NPVmethod, that is, the discounted sum of its futureincome stream, is the theoretically correctmethod for asset valuation, and it has beenrecommended by the revised SEEA. The incomestream is calculated as the net price, which isthe price of an asset minus the marginal costs ofextraction. In practice, net price is oftencalculated as price minus the average costs ofextraction because information about marginalcosts are unavailable, often leading to anupward bias in NPV.

It is best to calculate net rent from establishmentdata, but when the information is not available,aggregate data from the national accounts areused. Whatever the source of data, it isnecessary to estimate two components of costincluded in the operating surplus, or mixedincome part of value-added. The first is the costof capital, or so-called “normal profit,” which isusually viewed as either the cost of borrowingcapital or the opportunity cost of capital. Thesecond component is the earnings of the self-employed. This is essentially a payment forlabor which is not included in compensation ofemployees because, as the owners of business,the self-employed do not pay themselves anexplicit wage.

The NPV method of valuation requiresassumptions about future prices and costs ofextraction, about the rate of extraction, andabout the discount rate. It is often assumed thatnet price and level of extraction remainconstant, although when information is knownabout planned extraction paths, or expectedfuture prices, this information can beincorporated. A wide range of discount rateshave been used by different countries.

In much of the early work on environmentalaccounting (e.g., Repetto and others 1987,

Bartelmus and others 1992, Van Tongeren andothers 1993, UN 1993), the net price method wasused to value assets rather than NPV. The netprice method simply applies the net price in agiven year to the entire remaining stock. Basedon an interpretation of Hoteling’s Rule, it isequivalent to the NPV method under therestrictive assumption that the net priceincreases at the same rate as the discount rate.Although this assumption is unrealistic, the netprice method was widely used because itappeared to avoid the need to project future netprice or extraction paths. However, the methoddid not really avoid the need to make theseprojections, it simply made it unnecessary tomake them explicit.

The revised SEEA recommends NPV, and thismethod has come to be more widely used thanthe net price method in more recent work. Theonly significant exception has been the work onforest assets by Eurostat (2000b), which usedseveral methods, including variation of NPVand net price. NPC is used by Eurostat forvaluation of subsoil assets (2000a).

Regardless of the method, most asset valuationhas focused on the dominant commercial use ofa resource. Some assets may have multiple uses.For example, forests, in addition to providingtimber, may have a direct commercial use forthe recreation industry, as well as otherimportant but less direct uses, e.g., carbonsequestration, or watershed protection. Inprinciple, all these values should also beincluded in the value of the forest; in practice,they may not be included.

Depletion and depreciation

One of the major motivations for thepreparation of asset accounts has been toaccount for the depletion of natural capital. Thisis particularly important for resource-richcountries which may appear to perform wellaccording to conventional economic indicators,



but in fact are living off their (natural) capital ina manner that cannot be sustained indefinitely.The cost of depletion was initially measured asthe value (net price) of extraction of non-renewable resources, and, for renewableresources, the value of the volume of harvestabove sustainable yield.

It has since been recognized that this concept,derived from ecological concepts ofsustainability, is not consistent with theeconomic concept of depreciation used formanufactured assets in the SNA. (For furtherdiscussion, see Davis and Moore (2000) orVincent (1999)). El Serafy (1989) proposed onemethod of estimating depreciation, but it is notconsistent with the economic definition ofdepreciation and has not been widely used. Therevised SEEA proposes a measure of depletioncost more consistent with economicdepreciation: the change in the asset value fromone period to the next. However, severalalternative ways to measure this cost have beenproposed and no consensus has yet beenreached (Ryan 2000). As a result of theuncertainty over measurement of depreciationof natural capital, most countries do notmeasure it.

2.3 Pollution and material flow accounts

Pollution and material (including energy andresource) flow accounts track the use ofmaterials and energy and the generation ofpollution by each industry and final demandsector. The flows are linked through the use of acommon industrial and commodityclassification to IO tables and Social AccountingMatrices (SAMs), as exemplified by the DutchNAMEA framework, which has been adoptedby Eurostat and the revised SEEA manual.Much of the work on environmental accountshas been pioneered by industrialized countriesand reflects their major policy concerns.

Physical accounts

The most widely available accounts are forenergy and air emissions, especially emissions

linked to the use of fossil fuels. Energy accountshave been constructed by many countries sincethe dramatic oil price increases of the 1970s,and, since many air pollutants are linked toenergy use, it is relatively simple to extend theaccounts to include these pollutants.Transboundary flows of atmospheric pollutantsthat cause acid rain has been a major policyconcern throughout Europe for more than twodecades. More recently, the concern withclimate change has made tracking greenhousegas emissions a priority. Accounts are alsoconstructed for other air pollutants, waterpollutants, solid waste, and other forms ofenvironmental degradation such as soil erosion.In a growing number of countries, especiallywater-scarce countries, water accounts are ahigh priority (Australia, France, Spain, Chile,Moldova, Namibia, and Botswana).

Some studies have attempted to fully accountfor all environmental services including suchitems as watershed protection provided byforests and open space, pollination ofagricultural crops by wild bees, recreation, andaesthetic enjoyment of the environment(Nordhaus and Kokkelenberg 1999). This is verydifficult and not commonly undertaken.

Monetary accounts for environmentaldegradation

In many countries, assigning an economic valueto environmental benefits and damage may beconsidered the most effective way to influencepolicy, if not the most efficient way to designpolicy. However, there remains controversy overwhether these monetary estimates are properlypart of the environmental accounts or a separateanalysis of the (physical) accounts. Mostcountries attempt some valuation, even ifoutside what they define as the environmentalaccounts, using one (or sometimes both, forcomparison) of two different approaches tovaluation:



• Maintenance, or avoidance, cost approach,which measures the cost of measures toreduce pollution to a given standard

• Damage cost approach, which measures theactual damage caused by pollution, in termsof, for example, reduced agriculturalproductivity due to soil erosion, increasedcorrosion of structures from acid rain, ordamage to human health from waterpollution.

In addition, the willingness-to-pay methodologycan be applied to environmental degradation,although it is not widely used at this time.

Measuring maintenance cost requires setting alevel of acceptable emissions (which may bezero) and calculating the cost of introducingtechnology to reduce current emissions to thatlevel. It is often, though not always, assumedthat the end-of-pipe, pollution abatementtechnology would be used, rather than aredesign of industrial processes to preventpollution. For example, the avoidance cost ofpollution from motor vehicles often assumes theuse of catalytic converters added on to vehicles,rather than policies to reduce use of motorvehicles, such as subsidies for mass transit. Thepollution abatement approach is attractive forseveral reasons: it is clearly consistent with thepolluter pays principle; in the past, abatementtechnology dominated technological solutions;and in many instances, abatement cost isrelatively easy to measure than otherapproaches. In the above example, it is mucheasier to estimate the cost of widespread use ofcatalytic converters than to estimate thenecessary mass transit subsidies and thedevelopment of the corresponding mass transitinfrastructure.

Industrialized countries have a relatively longexperience with pollution abatement, so thatthey can estimate the costs of reducingpollution. Developing countries do not havesuch extensive experience and often “borrow”

expenditure data from other countries (amethod called benefits transfer) to calculatecoefficients for expenditure on pollution controlper unit of output. These are then used as acrude estimate of the expenditures that wouldbe required in order to reduce pollution in thecountry in question.

Calculating damages caused by pollution ismuch more difficult. Damages include loss ofagricultural productivity or productivity ofother resources, accelerated corrosion tostructures, and damages to human health.Although it is theoretically the best method, ithas not been used as often as the maintenancecost approach.

Monetary accounts for non-marketed resourcesValuation issues discussed in the SEEA havelargely focused on environmental degradation,but other non-market goods and services alsoneed to be valued. The use of near-marketgoods like non-market firewood or wild foodproducts are, in principle, included in the SNA,and many countries have included someestimate of these resources in the conventionalnational accounts. Water, on the other hand, isan example of an economically importantresource that is often either unpriced, or pricedin a way that is not related to its true economicvalue. Water valuation can be quite difficult andeven in the revised SEEA little guidance hasbeen offered.

2.4 Environmental protection and resourcemanagement accounts

This third component of the SEEA differs fromthe others in that it doesn’t add any newinformation to the national accounts butreorganizes expenditures in the conventionalSNA that are closely related to environmentalprotection and resource management. Thepurpose is to make these expenditures moreexplicit, and thus, more useful for policyanalysis. In this sense, they are similar to other



satellite accounts, such as transportation ortourism accounts, which do not necessarily addnew information, but reorganize existinginformation. This set of accounts has three quitedistinct components:• Expenditures for environmental protection

and resource management, by public andprivate sectors

• The activities of industries that provideenvironmental protection services

• Eenvironmental and resource taxes/subsidies.

The United States pioneered the collection ofenvironmental protection expenditure (EPE)data in 1972, but has curtailed this datacollection effort in the mid-1990’s. TheEuropean Union (European System forEnvironmental Expenditure Information, orSERIEE) and the OECD (Pollution Abatementand Control Expenditure system, or PACE) havecompiled environmental protection expenditureaccounts throughout the 1990s. These data aregenerally obtained from industry surveys.

Expenditure for environmental protectionrepresent part of society’s effort to prevent or toreduce pressures on the environment. However,the interpretation of indicators from the EPEaccounts can be ambiguous. The EPE conceptworks best for end-of-pipe, pollution abatementtechnologies in which an additional productioncost is incurred to reduce pollution. However,the growing trend in pollution managementstresses pollution prevention through redesignof industrial processes rather than end-of-pipetechnology. New technology may be introduced,perhaps during the normal course ofreplacement and expansion of capacity or forsome other reason, that reduces pollution. Thereis currently no way to estimate how much of thecost of this new technology, if any, should beattributed to environmental protectionexpenditures. In some instances, process-integrated measures that reduce pollution may

reduce costs and pollution simultaneously. TheEU is responding to this problem by beginningto collect data about the use of integrated-process technologies. Surveys of recycling arealso included.

These problems make spending on EPEextremely hard to interpret and, therefore, ofmore limited policy use. An increasing ordecreasing EPE cannot be interpretedunambiguously as either a positive or negativetrend. If EPE accounts are constructed at thefirm level and tied to physical flow accounts atthe firm level, then one can one interpret thedata, and even then, only with the help ofadditional information to determine whetheradditional measures not included in EPEaccounts were also undertaken.

2.5 Macroeconomic indicators

The three sets of account described above eachprovide a range of indicators, but, with theexception of the asset accounts, these indicatorsdo not directly affect the conventionalmacroeconomic indicators such as GDP andNDP. Many practitioners have searched for away to measure sustainability by revisingconventional macroeconomic indicators or byproducing alternative macro indicators inphysical units. Table 2 lists the majorenvironmental macroeconomic indicators.

Physical indicators

Macroeconomic indicators measured in physicalunits have been proposed either as analternative to monetary indicators, or to be usedin conjunction with monetary aggregates inassessing economic performance. Physicalindicators reflect a strong sustainabilityapproach. The two major sources of physicalmacroeconomic indicators are the NAMEAcomponent of the SEEA flow accounts andMaterial Flow Accounts, which are closelyrelated to environmental accounts (see section



Table 2. Environmental macroeconomic indicators, physical and monetary

Source: Part B adapted from Table 1, Chapter VIII of the revised SEEA (UN and others 2000).

A. Physical Aggregates

Aggregates Title Basis

Physical macroeconomic aggregates

NAMEA Theme

Indicators, No acronym

Theme indicators for

greenhouse gas emissions,

acidification, eutrophication,

and solid waste

Derived from the NAMEA system, embedded

in the SEEA flow accounts

TMR, DMI, NAS, TDO,

DPO

Total Material

Requirements, Direct

Material Input, Net

Additions to Stock, Total

Domestic Output,

Domestic Processed Output

Derived from Material Flow Accounts

B. Monetary Aggregates

Aggregates Title Basis

1. Measures that revise existing macroeconomic indicators

daGDP, daNDP, daGNI,

daNNI

Depletion adjusted product

and income measures

Subtract depletion of natural capital assets

from macroeconomic aggregates

eaNDP, eaNNI “Environmentally adjusted”

product and income

Subtract depletion of natural capital and

environmental degradation based on

maintenance cost from macroeconomic

aggregates

In some implementations, part of EPE are

subtracted.

Genuine income (gY) NNI less damage costs;

(related to genuine savings,

goods and services)

Subtract depletion of natural capital and

environmental degradation based on damage

cost from macroeconomic aggregates

Genuine Savings, no

acronym

Genuine Savings Revise conventional measure of Savings for

net change in natural capital and human capital

2. Measures that estimate new, hypothetical macroeconomic aggregates

Hueting's Sustainable

national income (SNI)

Sustainable income measure

preserving environmental

services

Modeling of hypothetical GDP, GNI if

economy was forced to meet environmental

standards using currently available technology

geGDP, geNDP, geGNI,

geNNI

“Greened economy”

product and income

measures,

Modeling of hypothetical GDP if hypothetical

environmental protection costs were required

Other forms of

sustainable GDP, NDP,

GNI, NNI

No technical term Modeling of hypothetical GDP from a range of

either short- and medium-term options (e.g.,

carbon tax) to long term strategic analysis of

alternatives for sustainable development



4.1.B for further discussion of MFA and theSEEA).

The NAMEA provides physical macroeconomicindicators for major environmental policythemes: climate change, acidification of theatmosphere, eutrophication of water bodies,and solid waste. These indicators are compiledby aggregating related emissions using somecommon measurement unit, such as carbondioxide equivalents for greenhouse gases. Theindicators are then compared to a nationalstandard—such as the target level ofgreenhouse gas emissions—to assesssustainability. The NAMEA does not, however,provide a single-valued indicator whichaggregates across all themes.

The Material Flow accounts (MFA) provideseveral macro indicators; the most widelyknown is total material requirements (TMR) (seeBartelmus and Vesper (2000), World ResourcesInstitute (2000)). TMR sums all the material usein an economy by weight, including so-called“hidden flows,” which consist of materialsexcavated or disturbed along with the desiredmaterial, but which do not themselves enter theeconomy. In contrast to NAMEA themeindicators, TMR provides a single-valuedindicator for all material use. Sustainability isassessed in terms of “dematerialization” such asFactor 4 (halving resource use while doublingwealth (von Weisäcker and others 1997).However, TMR does not differentiate materialsby their environmental impact—highly toxicmaterials are simply added to materials liketimber or gravel that may be much lessenvironmentally damaging. Consequently, thesustainability goals set under this frameworkappear rather vague to be used as guides topolicy, and require more disaggregation, bymaterial and by industry, to be interpretedcorrectly. See Cleveland and Ruth (1999) for acriticism of these physical indicators ofdematerialization.

Monetary indicators

The purpose of most monetary environmentalmacroeconomic aggregates has been to providea more accurate measure of sustainable income.The first approach revised conventionalmacroeconomic indicators by adding andsubtracting the relevant environmentalcomponents from the SEEA, the depletion ofnatural capital (daNDP) and environmentaldegradation (eaNDP) (O’Connor 2000). Theadjustment of NDP for asset depletion (daNDP)is accepted in principle by most economists andstatisticians, even though there is not yet aconsensus over the correct way to measure it.Environmentally-adjusted NDP has beencriticized for combining actual transactions(conventional NDP) with hypothetical values(monetary value of environmental degradation).If the costs of environmental mitigation hadactually been paid, relative prices throughoutthe economy would have changed, therebyaffecting economic behavior and, ultimately, thelevel and structure of GDP and NDP.

A macro indicator related to eaNDP is GenuineSavings, reported in the World Bank’s WorldDevelopment Indicators (Kunte and others 198, Hamilton 2000, World Bank 1999). It attempt to measure changes in asset values rather thanincome. According to economic theory, (weak)sustainability requires that wealth is non-declining over time. Many countries do nothave comprehensive balance sheets, so it is notyet feasible for them to monitor wealth.However, it is possible to measure savings moreaccurately, which indicates how wealth ischanging, and whether the trend is sustainableor not. Genuine savings adjusts gross domesticsavings for consumption of fixed capital,investment in human capital, changes in naturalcapital, and environmental damage.

The criticism of eaNDP led to the constructionof a second approach to constructing indicators,which asked the question, what would GDP or



NDP have been is the economy were required tomeet sustainability standards. These indicatorsof a hypothetical economy are derived througheconomic modeling. Two modeling approacheswere developed:• Hueting’s Sustainable National Income

(SNI) which estimates what the level ofnational income would be if the economymet all environmental standards usingcurrently-available technology (Verbruggenand others 2000)

• geNDP which estimates how the economywould respond if the estimatedmaintenance costs were internalized in theeconomy.

Hueting’s SNI is the maximum income that canbe sustained without technologicaldevelopment (excluding the use of non-renewable resources). It is not meant torepresent what the economy should look like,but rather, to show to policy-makers thedistance between the current economy and asustainable economy. The geNDP approach is a

bit less restrictive, technological change ispossible, depending on the time period for thetransition. The purpose of this approach is toprovide policy-makers with guidance about theenvironmental impacts of alternativedevelopment paths and the most efficient policyinstruments for meeting environmentalobjectives.

Identifying a sustainable national income is ahighly complex undertaking. It requireseconomic modeling that includes assumptionsabout the environmental standards to achieve,the technological means to achieve them, theresponse to policy instruments, and the usualrange of assumptions for an economic model:income and price elasticities, impacts on trade,etc. Different combinations of these options andassumptions about the future will result in quitedifferent levels of sustainable national income.Much depends on the period of time over whichsustainable income would be achieved. Becauseof this complexity, no studies have producedindicators that are comparable across countries.


Asset Accounts3One of the fundamental indicators of acountry’s well being is the value of its wealthover time. The discussion of sustainabilityindicated that there are different views abouthow wealth should be measured, i.e., if allforms of wealth can be measured in monetaryterms (weak sustainability) or if wealth must bemeasured in some combination of monetary andphysical units (strong sustainability). Whetherone chooses to aggregate different forms ofwealth, and whether one interprets theaggregate figure as an indicator of sustainabilityor not, it is certainly necessary for a country tomonitor its wealth over time. While non-declining national wealth does not guaranteesustainable development, declining nationalwealth almost certainly a cause for concern.Better, more comprehensive accounts fornational wealth can only improve the ability ofresearchers and policy-makers to makeinformed decisions.

This section begins with a review of the wayasset accounts contribute to more effectivemonitoring of national wealth, then discusseshow the asset accounts can be used to improvemanagement of natural capital. Some of theissues that are addressed include:

Monitoring national wealth• Physical stocks of natural capital and the total

economic value of produced and non-produced assets

• Change in per capita wealth over time• Depletion of natural capital and the economic

cost of depletion.

Analysis of the management of national wealth• · Is the resource rent being recovered

successfully by government througheconomic instruments

• Is rent used to promote a sustainableeconomy—e.g., rent from non-renewableresources reinvested in other activities thatcan take their place

• Is the maximum rent being generated bynatural resource policies

• If not, are there other socio-economic objectivesthat are being met, such as support to ruraleconomies, or employment creation, andwhat is the economic cost of meeting theseother objectives.

3.1 Monitoring total wealth and changesin natural capital

The asset accounts provide fundamentalindicators to monitor sustainability: the value ofwealth and how it changes from one period tothe next through depreciation or accumulation.Although total wealth and per capita wealth,expanded to include both manufactured andnatural assets, are useful indicators, not manycountries compile such figures yet. Instead,many countries have focused on compilingaccounts for individual resources andsometimes estimating depletion of naturalcapital, which is used to compile a morecomprehensive measure of depreciation than isfound in the conventional national accounts.This section begins with a review of the use ofthe asset accounts to measure wealth, and thenturns to measures of depletion and depreciation.



PHYSICAL ASSET ACCOUNTS

The physical asset accounts provide indicatorsof ecological sustainability and detailedinformation for the management of resources.The volume of mineral reserves, for example, isneeded to plan extraction paths and indicateshow long a country can rely on its minerals. Thevolume of fish or forestry biomass, especiallywhen disaggregated by age class, helps todetermine sustainable yields and the harvestingpolicies appropriate to that yield.

The asset accounts track the changes in stockover time and indicate whether depletion isoccurring. They can, thus, show the effects ofresource policy on the stock and can be used tomotivate a change in policy. For example, thebiological depletion of Namibia’s fish stockssince the 1960’s has provided a very clearpicture to policy-makers of the devastatingimpact of uncontrolled, open-access fishing(Figure 1). Similar accounts of depletion (oraccumulation) have been constructed for forestsin the Philippines, Brazil, Chile, Malaysia,Indonesia, Australia, Canada and much of theEU.

Physical accounts for land accounts track theuse of land for different purposes, and itsconversion from one use to another over time,which can be linked to the environmental andeconomic consequences of conversion, such asincreased soil erosion or loss of watershedprotection. For example, a major concern indeveloping as well as industrialized countries,concerns the conversion of high potentialagricultural land to urban settlements, the lossor fragmentation of fragile ecosystemssupporting unique biodiversity, or the clearingof forests for agriculture. Land accounts andindicators of environmental pressure related toeconomic activity have been constructed in anumber of industrialized countries, such asCanada (1997, 2000), Germany (Rademacher1998), and the UK (Stott and Haines-Young1998) but have not been widely used indeveloping countries. Limited accounts wereconstructed for Korea. The Philippinesconstructed a model to assess the economicconsequences of different patterns of land use(ENRAP 1998). Where land has been addressed,it is usually in relation to soil and, moreparticularly, degradation of soil through erosionor other factors.

Some environmental assets clearlyyield economic benefits, which maynot always be fully captured inmonetary terms such as carbonsequestration, watershed protection,or provision of habitat to maintainbiodiversity. For such assets, physicalaccounts often provide the bestalternative for management. Forexample, physical accounts for stocksand flows of greenhouse gases (GHG)are required to address climatechange. An important element inthese accounts is the carbon storagecapacity of natural assets, especiallyforests. Carbon binding is calculatedas a given percentage of the estimated

Figure 1. Biomass of hake, pilchard, horse mackerel in

Namibia, 1963 to 1999 (thousands of tons)

Source: Lange 2001.

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

1963

1965

1967

1969

1971

1973

1975

1977

1979

1981

1983

1985

1987

1989

1991

1993

1995

1997

1999

Horse Mackerel

Hake

Pilchard


Asset Accounts

biomass of forests, and changes in carbon stocksare estimated on the basis of changes in forestbiomass. Such accounts are constructed forAustralia and Canada, and were constructed forSouth Africa’s forests in a recent academic study(Hassan 2000).

International protocols for climate change whichallow tradeoffs between carbon emissions andcarbon sinks, such as forests, will make theseaccounts essential. Limited international tradingin carbon sinks is already occurring. If this trendexpands, it will be very useful for developingcountries with large forest potential to compileforest and carbon accounts.

MONETARY ASSET ACCOUNTS

The physical accounts for individual assets canbe used to monitor ecological sustainability.However, a more complete assessment of assetsrequires that the economic value of a resourcealso be known. The monetary value of differentassets, produced and non-produced, can becombined to provide a figure for total nationalwealth. This figure can be analyzed to assess thediversity of wealth, its ownership distribution,and its volatility due to price fluctuations, animportant feature for economies dependent onprimary commodities. Diversity is importantbecause, in general, the more diverse aneconomy is, the more resilient it will be toeconomic disturbances. Volatility is alsoimportant in planning for the future—lowervolatility contributes to more stable economicdevelopment. The distribution of the ownershipof assets—between public and private sector,the concentration among different groups insociety, and between domestic and foreign—canhave significant economic implications and caninfluence the sustainable management ofresources.

Most countries with asset accounts for naturalcapital have typically published the accountsseparately for each resource and have notattempted to measure of total natural capital

(the sum of all resources), or a measure of totalnational wealth (the sum of manufactured andnatural capital). Among the developingcountries, Botswana (Lange 2000a) and Namibia(Lange 2001) are doing so. Among theindustrialized countries, Australia (ABS 1999)and Canada (Statistics Canada 2000) haveintegrated non-produced natural assets withproduced assets in their Balance Sheets.

In some cases, this may simply reflect a relativelack of concern by policy-makers aboutwealth—most countries have traditionally beenmuch more concerned with the income andproduct flow accounts of the SNA than with theasset accounts. Some developing countries, likethe Philippines, may not have data aboutmanufactured capital stock. In other cases, theremay also be a reluctance to combineconventional measures of manufactured capitalwith what may be viewed as experimentalcalculations for natural capital, especially whenthere is controversy over the assumptionsnecessary for valuation (Ryan, 2000), or over thepolicy implications of the results, e.g., fisheriesin Chile.

In the discussion that follows, the total wealthof three countries are compared—Australia,Botswana, and Namibia. Although accounts fornatural capital are not comparable across thesecountries because of differences in methodologyand coverage, international comparisons arenonetheless instructive, and provide the basisfor improving the accounts.

Australia’s accounts for natural capital includeland, subsoil assets, and native forests in non-produced asset accounts, and commercialforests in produced assets (Figure 2.A).Botswana’s accounts currently include onlymineral assets (Figure 2.B), and Namibia’s assetaccounts include minerals and fisheries (Figure2.C). Neither Botswana or Namibia have valuedland yet because no market price exist for thevery large portions of the land that are subjectto traditional communal tenure regimes. Other



Figure 2. Value of produced and non-produced assets in Australia, Botswana and Namibia in current

prices

2.B Botswana

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97

Mineral assets

Prod. Assets, private

Prod. Assets, govt.

2.C Namibia

10,000

20,000

30,000

40,000

50,000

60,000

Millio

ns

of

Nam

ibia

do

llars

70,000

1990 1991 1992 1993 1994 1995 1996 1997 1998

Fisheries

Minerals

Produced capital, govt.

Produced capital, private

0

Source: Australia—Australian Bureau of Statistics 2000, Botswana—Lange 2000, Namibia—Lange 2002.

2.A Australia, 1989 to 1998

0

500

1000

1500

2000

2500

3000

Billio

ns

of

Au

stra

lian

do

llars

1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Native forests

Subsoil assets

Land

Produced assets

significant missing assets include wildlife forboth Botswana and Namibia, and water for allthree countries. Figures are reported in currentprices because there is not yet an agreed uponmethodology for constant price calculations fornatural capital.

The trend of national wealth over timeindicates, at an aggregate level, whether capitalis maintained, whether depreciation of naturalcapital, if it has occurred, has been compensatedfor by an increase in other forms of capital. In all

three countries the value of total capital hasincreased over time. The value of Australia’stotal (non-financial) capital increased 43percent, and roughly doubled in Botswana andNamibia during the 1990’s, despite exploitationof non-renewable resources. Whatever depletionhas occurred in these countries has beencompensated for by a combination of factors:new discoveries, an increase in economicallyprofitable reserves, and an increase in theeconomic value of reserves. In the case ofNamibia’s fisheries, improvements in fisheries


Asset Accounts

management has also played a role in theincreased value of the fish stock.

In addition to the volume of wealth, thecomposition of wealth is also an importantindicator to monitor because, generally, a morediverse economy is a more resilient one. Manyresource-rich developing countries haveidentified economic diversification as one of theirdevelopment objectives. A comparison of theshares of produced and natural capital over timeis one approach to monitoring this aspect ofresilience and progress toward diversification.

Natural capital is quite important in all threecountries (Table 3). Botswana is the onlycountry where the share of natural capital, 52%in 1996, is greater than manufactured capital,48%. This share is significantly lower than its65% share in 1990, which indicates substitutionof produced capital for mineral capital overtime, and diversification of the economy. For anindustrialized country, Australia has a relativelyhigh dependence on natural capital—37% of itsnon-financial assets in 1998, but not muchdifferent from 1990. The share of natural capitalis much smaller in Namibia, only 18% by 1998,but higher than its share in 1990 (14%). In theAustralian accounts, land dominates naturalcapital, accounting for roughly three-quarters ofthe total value of non-produced assets. NeitherBotswana or Namibia have included land intheir asset accounts, probably resulting inserious underestimates of national wealth.

Australia’s national wealth shows relatively lowvolatility, possibly because land is the largestcomponent and land value is not subject to thelarge changes in price and demand faced byglobally-traded resources like minerals and fish.Although Botswana has seen a dramatic changein the relative shares of produced and non-produced capital, the transition has beensmooth, without major changes from year toyear. By contrast, the natural assets of Namibiaappear to be less stable, accounting for 14% ofwealth in 1990, falling to a low of 8% in 1996,followed by a high of 18% in 1998. This trendthat is more pronounced when its natural assetsare examined individually.

Namibia’s mineral assets consist primarily ofdiamonds, uranium and gold. The combinedvalue of these assets (confidentiality concernsprevents reporting the value for each mineral)fell from N$2,575 million in 1990 to N$976million in 1993, then recovered steadily to justover its 1990 value in 1998. Under a newmanagement system since independence in1990, Namibia’s fisheries are recovering from along period of over-exploitation, which makesthe stock extremely vulnerable to disturbancesof the marine environment. The value of the fishstock fell and recovered twice during the 1990’s,rising dramatically by more than 100% between1997 and 1998, due partly to fluctuations in thetock and partly to fluctuating economicconditions. Clearly, Namibia’s wealth is muchmore volatile than the other two countries.

Table 3. Natural capital as percent of total non-financial assets in Namibia, Botswana,

and Australia, 1990 to 1998 (percent)

Notes: Na: not available.

Figures for Australia include only non-produced assets, not types of natural capital which are included under produced

capital.

Percentage shares calculated in current prices.

Source: Derived from: Australia—Australian Bureau of Statistics 2000, Botswana—Lange 2000. Namibia—Lange 2002.

1990 1991 1992 1993 1994 1995 1996 1997 1998

Namibia 14 12 10 10 12 11 8 11 18

Botswana 65 59 57 53 52 54 52 na na

Australia 34 34 32 33 34 35 34 36 37



In using national wealth to monitor economicsustainability, it is crucial to include all assets,or at least as many as possible (human capital isnot yet included). Some natural assets are notyet included in the countries accounts, mainlybecause of valuation difficulties. Anotherimportant component of a country’s assetportfolio is its net foreign financial assets. It isnot uncommon for resource-rich developingcountries, like some of the major oil-producingcountries, to invest much of the income fromresource exploitation in foreign assets,especially if the economy is small andopportunities for domestic investment arelimited. For such countries, net financial assetsform a significant share of national wealth.Indeed, for Botswana, foreign financial assetshave grown increasingly important, from 13% oftotal net worth in 1990 to 21% in 1996. In thecase of Australia, net foreign financial assets arenegative and growing, reaching –16% of networth in 1998 (Table 4). Figures are not yetavailable for Namibia.

So far, we have considered only trends in totalwealth. However, in most countries, population

is still increasing, so a constant level of wealthand income would result in a declining percapita level of wealth and income for futuregenerations. Inter-generational equity requiresthat not just total wealth, but per capita nationalwealth is non-declining over time.

Figure 3 shows the index of per capita wealth incurrent prices from 1990 to 1998 (1996 forBotswana). For Botswana and Australia, the networth figure including net foreign financialassets is used; for Namibia, only produced plusnatural capital are available. Per capita wealthhas grown the fastest for Botswana, by 80%from 1990 to 1996, and the slowest for Australia,only by 22%, which is not surprising for amature industrialized economy. Per capitawealth also grew fast in Namibia, though not asfast as Botswana, though without informationabout foreign financial assets, it is not certainwhether the figure reported understates oroverstates asset growth. The trends for the threecountries are not readily comparable becausethey are not in constant price terms.

Table 4. Financial, non-financial assets, and net worth in Botswana and Australia,

1990 to 1999

Na: not available

Notes: Figures are in current prices. Figures for Namibia are not currently available.

Source: derived from: Australia: Australian Bureau of Statistics, 2000. Botswana: Lange, 2000.

1990 1991 1992 1993 1994 1995 1996 1997 1998

Botswana,millions of pula

Produced + Non-

produced assets 35,657 37,947 40,147 43,764 50,734 58,351 66,613 Na Na

Net foreign financial

assets 5,482 6,919 7,595 9,413 10,693 11,871 17,636 19,378 24,517

Net worth 41,139 44,866 47,742 53,177 61,427 70,222 84,249 Na na

Australia, billions of A$

Produced + Non-

produced assets 1,742 1,796 1,797 1,885 1,984 2,095 2,154 2,285 2,428

Net foreign financial

assets -171 -192 -201 -217 -238 -263 -279 -303 -321

Net worth 1,572 1,604 1,596 1,668 1,746 1,832 1,875 1,982 2,107


Asset Accounts

The public and private sectors may havedifferent resource management objectives whichaffect the way resources are exploited.Consequently, monitoring the distribution ofasset ownership between public and privatesectors may be useful, not as a direct indicatorof sustainability, but as an aid to resourcemanagement. The private sector is motivatedlargely by commercial concerns, which canfavor economic efficiency, but also depletion ofrenewable resources under certain conditions.Government may or may not utilize resources ina sustainable, and it may use resources toachieve other socio-economic objectives, even ifthis lowers the economic return from a resource.

The response to the depletion of natural capitalmay also differ between the private and publicsector. Where depletion occurs, sustainabilityrequires reinvestment in other forms of capital.Private ownership may result in reinvestment inprivate sector activities, but foreign ownershipmay result in reinvestment elsewhere which

does not benefit the country providingthe wealth. In countries where thegovernment owns the resource andrecovers most of the resource rent, thegovernment bears responsibility forreinvestment, often investing in publicsector capital. There is disagreementover the extent to which growth ingovernment assets is an effectivesubstitute for other forms of capital.There is a tendency to assume thatgovernment is economically inefficientcompared to the private sector, but it isalso well documented that the privatesector will under-invest in assets wheresocial benefits exceed private benefits,like public infrastructure and humancapital. In any case, it is useful tomonitor the distribution of capital, aswell as its level and composition.

In Botswana and Namibia, mineralsand fisheries are owned by the state.

Private sector manufactured capital has beengrowing faster than public sector manufacturedcapital in Namibia, while the opposite is true inBotswana (Figure 2). In Australia, non-producedassets, dominated by land, are mainly (70%)owned by the private sector.

DEPLETION AND DEPRECIATION OR

ACCUMULATION OF NATURAL CAPITAL

In addition to a measure of overall wealth, assetaccounts may provide a measure of annualdepreciation of produced assets and naturalcapital. The depletion of natural capital reducesthe wealth available to future generations, andreduces the income-generating capacity of theeconomy. If this loss is not reflected in thenational accounts, policymakers mayoverestimate the level of sustainable nationalincome.

Depletion accounts, mostly using the net pricemethod rather than economic depreciation,

Notes: Index calculated in current prices.

National wealth is net worth for Australia and Botswana. For Namibia, only

produced + natural capital are included.

Source: derived from: Australia—Australian Bureau of Statistics 2000, Botswana—

Lange 2000, Namibia—Lange 2001.

Figure 3. Index of per capita national wealth in Australia,

Botswana, and Namibia, 1990 to 1998 (1990=1.00)

1.00

1.10

1.20

1.30

1.40

1.50

1.60

1.70

1.80

1.90

1990 1991 1992 1993 1994 1995 1996 1997 1998

Namibia

Botswana

Australia



were calculated in the early studies forIndonesia and Costa Rica by Repetto (1989), forMexico (van Tongeren and Schweinfest 1991)and Papua-New Guinea (Bartelmus and others1992) by the UN and World Bank, and in morerecent work undertaken by the Philippines(forests, minerals, land, fish), Chile (forests),Indonesia (forests, minerals) and Brazil (forests).Indonesia reported that depletion accounted for11% of GDP over the period 1993 to 1996 (Saleh2000). Estimates of depletion or depreciationhave been estimated in a number of one-time, oracademic studies such as Malaysia (Vincent1997).

Depletion estimates for the Philippines aredominated (90% or more) by the costs ofdeforestation, which was halted by a loggingban in 1992 (Table 5). Depletion for minerals isquite low, reflecting the slow-down in miningdue to falling world mineral prices from 1991 to1993. Depletion of land and fisheries have bothbeen increasing, fisheries at a rather alarmingrate.

Controversy over measurement has made itmore difficult to provide policy-makers with adefinite figure for depletion (See discussion insection 2) and many of the industrializedcountries have not calculated it, or areconducting experimental calculations, not fullyintegrated into the national accounts. The recentpublication by Eurostat of pilot accounts for

subsoil assets (Eurostat 2000a) and for forests(Eurostat 2000b) did not attempt to measure thecost of depletion. Australia used severalalternative methods for valuing the cost ofdepletion for sub-soil assets during the 1990’s.Including this depletion with conventionalmeasure of depreciation would reduce NetDomestic Product by less than 1%, which is notvery significant (Ryan 2000, Tables 2, 8).

Of course, not all change in the stock ofresources is negative. There has been a netincrease in the volume of cultivated timber inSouth Africa. Using the change in asset valueapproach of Vincent (1999), Hassan (2000)found this would increase Net National Productby roughly 0.5% over the past decade. Forcountries like Botswana, if the cost of depletionis calculated using the old method—net pricetimes the quantity extracted—then this cost willbe quite high. However, if the economic methodof depreciation, the change in asset value, isused, there is a net appreciation because ofholding gains.

3.2 Managing resources—economicefficiency, sustainability, and other socio-economic objectives

In the early days of environmental accounting,resource rent was calculated in order tocalculate the value of assets, but its usefulnessas a resource management tool was not alwaysrecognized. More recent work by Norway

Table 5. Cost of depletion of natural capital in the Philippines, 1988 to 1993

(millions of pesos)

Note: depletion cost estimated using the net price method.

Source: NSCB.

1988 1989 1990 1991 1992 1993

Forests 24,529 14,852 10,404 11,766 542 300

Land 918 979 1,212 1,618 1,401 1,241

Fisheries 188 366 573 936 1,106 1,073

Minerals 381 296 81 1 0 0

Total 28,004 18,482 14,260 16,312 5,041 4,607


Asset Accounts

(Sorenson and Hass 1998), Eurostat (2000a) forsubsoil assets, in the Philippines (ENRAP 2000,Lange 2000b), Botswana (Lange 2000a),Namibia (Lange and Motinga 1997, Lange2001b), and South Africa (Blignaut and others2000) has included detailed analysis of resourcerent. Rent has been used to assess resourcemanagement in terms of economic efficiency,sustainability, and achieving other socio-economic objectives, such as inter-generationalequity.

In terms of economic efficiency, some of theissues addressed include:• How much resource rent is being generated

and is it being recovered by government oraccruing to the private sector?

• Do payments to government by the privatesector at least cover the cost to governmentof managing the industry?

• Do current management policies maximizethe amount of rent that can be generatedfrom the resource, or could rent be higherunder an alternative management regime?

In terms of sustainability, issues include:• Is rent from nonrenewable resources being

reinvested?• Do pricing policies for renewable resources

promote sustainable management?

In terms of other socio-economic objectives,issues include:• Do policies contribute to inter-generational

equity?• How are the benefits from resource

exploitation shared among different groupsin society?

3.2.A Economic efficiency

The value of an asset depends, in part, on howefficiently it is exploited. Norway provides agood example of how analysis of rent can beused to compare very different approaches toresource management that affect the value of

three major natural assets—petroleum,uncultivated forests, and fisheries. Figures 4,5,and 6 show the rent generated by each of theseresources and the distribution of that rentbetween government and private sector. Inmany countries, resources belong, by law, to thestate. As the owner of the resource, thegovernment has a right to charge for theexploitation of the resource by privatecompanies, and the responsibility to ensureproper use of the rent. Often, resource pricingdoes not recover the full rent, or even the fullcost of managing the resource.

Significant amounts of rent have been generatedby the oil and natural gas industry of Norway,but the rent has fluctuated a great deal, evenbecoming negative in 1988, meaning that, inthat year, the industry did not even earn anormal return to the capital it had invested inthe industry. To understand the high volatilitybetter, one must distinguish between total rentand per unit rent (not shown) to determine howmuch of the rent fluctuation results fromchanges in the volume of extraction, and howmuch results form changes in market price orthe cost of production. The figures forpetroleum indicate that governmentappropriates most of this rent through taxes. In1986 to 1988, the taxes collected by governmentactually exceeded rent.

Uncultivated forests are also managed in a waythat generates substantial rents, but, rather thancontributing this rent to government, the rentaccrues to the private sector which receives, inaddition, large subsidies (Figure 9). Fisheries(Figure 10) provide an example of yet a thirdmanagement strategy. Fisheries (these figuresinclude both capture fisheries and aquaculture)are managed in a way that generates no positiverent—rent from fisheries has been negativeexcept in 1995, and, rather than collectingrevenue, government has been providing theindustry with considerable subsidies. Themanagement regime promotes exploitation of



-10,000

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996

Resource rent

Government appropriation

mill

ions

of

NO

KFigure 4. Resource rent and taxes from oil and gas mining in Norway,

1985 to 1996

Figure 5. Resource rent and taxes from forestry in Norway, 1985 to 1995

-400

-200

0

200

400

600

800

1,000

1,200

1,400

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

Rent

Govermentappropriation

mill

ions

of

NO

K


Asset Accounts

fish stocks by relatively small inefficient vesselsin order to support Norway’s regionaleconomies. The economic value of fish underthis management regime is zero.

A pilot study by Eurostat for oil and natural gasfound that in the Netherlands, governmentappropriated between 82% and 97% of the rentbetween 1990 and 1998. The figures weregenerally lower for the UK; after the industrybegan to earn a positive rent in 1993,government’s share ranged from 45% to 99%(Eurostat 2000a). Rent recovery in thePhilippines was generally quite low for allresources except minerals. The studies ofNamibia, Botswana, and South Africa (sub-soilassets and, in Namibia, fisheries) showed mixedresults: rent recovery from sub-soil assets wasquite high for Botswana and Namibia, but lowfor Namibian fisheries. Rent recovery in SouthAfrica was mixed: high from gold mining butlow from coal.

These figures can provide a basis for settingresource pricing policy. The figures for Namibiawere one of several factors prompting a review

of quota fees for fisheries. In the Philippines, therent for grasslands was used in negotiationsbetween government and ranchers to revisegrazing fees (Batcagan 2000).

In assessing the contribution of the resource tothe economy, it is also useful to calculate thecontribution of rent taxes to total governmentrevenues, and the share of resourcemanagement costs incurred by government thatare recovered through taxes on rent. These usesof SEEA for resource management are fairlyrecent recommendations in the SEEA and havenot been carried out routinely by all countriescompiling asset accounts.

Government’s reliance on tax revenues from itsresources may be quite significant in some small,resource-dependent economies. For example, thegovernment of Botswana receives about 50% ofits revenue from taxes on mining. Even inNorway, the petroleum industry makes asignificant contribution to government revenue,7–8% in recent years (Central Bureau of Statistics,unpublished data 2000). A comparison of rentand resource management costs has been carried

Figure 6. Resource rent and subsidies to fisheries in Norway, 1985 to 1995

Source: Norwegian Central Bureau of Statistics 2000.

-4,500

-4,000

-3,500

-3,000

-2,500

-2,000

-1,500

-1,000

-500

0

500

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995

Rent

Govt. appropriation

mill

ions

of

NO

K



out by some European countries, as well as Chile,Costa Rica, and Namibia.

ECONOMIC EFFICIENCY — POTENTIAL VS. ACTUAL

VALUE OF ASSETS

Resource management can be evaluated fromthe point of view of economic efficiency todetermine if alternative policies might increasethe income generated and, hence, the economicvalue of a resource. While it is unlikely that anyindustry is perfectly efficient, the fisheriesindustry in Norway represents an extreme caseof economic inefficiency where the economicvalue of its fish stocks is zero.

Analysis of micro-survey data of Norway’sherring fishery found significant differences inrent-earning capacity between large and smallfishing vessels. Generally, the large-scaleoperations were more efficient and generatedrent. Assuming that most of the fishery couldpotentially be managed in a such an efficientmanner, one study of Norway’s herring fisheryestimated the potential resource rent of 1,000million Norwegian kroner (Flam 1993 quoted inSorensen and Hass 1998). A similar observationwas made of Namibia’s hake fishery, where therent generated by the three most profitablefishing companies earned rents more thandouble the industry average. The data were notsufficient to support calculation of potentialrents, but large differences in efficiency exist,mostly attributable to the differences betweenlarge, experienced operators with goodinternational connections, and the smallerNamibian newcomers to the industry (Lange2001b).

EFFICIENCY WITH MULTIPLE USES OF A RESOURCE

The values for assets discussed so far have beenbased on a single use; Norwegian forest assets,for example, consider only the timber value.The full economic value of an asset and theefficient management of the asset must be basedon an accounting of the full range of

environmental services which the asset canprovide. Forests may provide multiple benefits,such as timber, recreational benefits, carbonsequestration, and the provision of traditionalmedicines and foods, which can be importantfor rural populations in developing countries. Inpractice, comprehensive valuation of the may bedifficult to achieve.

Although the non-timber benefits are moredifficult to value, it is increasingly important todo so. The carbon sequestration value of forestswill be increasingly important as progress ismade on an international agreement to addressclimate change. Physical accounts for carbonstorage by forests were discussed earlier, butvalues for carbon storage have not beensystematically included in forest accounts.Carbon sequestration rights of tropical forestshave already been purchased or rented by someelectric utility companies in North America tooffset the companies’ carbon emissions. If acondition of this purchase or rental is thattimber not be cut, or that logging occurs on amuch reduced basis, then valuing such a forestat timber value alone may underestimate itstrue economic value.

Similarly, forests in Alaska have providedsubstantial economic benefits to the logging,recreation, and fishing industries. Economicefficiency requires assessing the optimal mix ofthese competing uses. Valuation of forests basedon timber alone would underestimate the realeconomic value of the forests.

3.2.B Sustainability

Economic sustainability requires that a portionof the rents from non-renewable resources (orfrom unsustainable exploitation of renewableresources) be re-invested in other assets oreconomic activities. In this way, exploitation ofthe resource can be economically sustainable—because it creates a permanent source ofincome—even though non-renewable resources


Asset Accounts

are, by definition, not biologically sustainable.For renewable resources, like forests or fisheries,one of the economic instruments formanagement includes resource pricing or userfees. To encourage sustainability, fees should beset high enough to recover the rent generated atthe most profitable, sustainable level ofproduction, so that it becomes unprofitable forcompanies to harvest at levels that deplete theresource stock.

Policymakers have sought indicators of theshare of rent from non-renewable resources thatshould be reinvested to maintain sustainability.Hartwick’s rule seemed to provide someguidance, stating that all net returns fromexhaustible resources should be reinvested(Hartwick 1997), but this now appears to bemore a descriptive rule than a prescriptive one(Asheim and Buchholz 2000). El Serafy providesanother approach. The SEEA does not provideguidance on this issue and no countries have sofar integrated this concept into the sustainabilityindicators they compile, with, perhaps, theexception of Botswana.

Botswana, whose economy is highly dependenton mineral revenues (roughly 35% of GDP), hasdeveloped the Sustainable Budget Index (SBI) toindicate how much of the mineral revenues areused for capital expenditures includingspending for human capital on education andhealth (Ministry of Finance and DevelopmentPlanning 1997, Wright 1997). Although there areno strict rules for policy based on the SBI,government has adopted an informal fiscalguideline that no revenues from mining shouldbe used for recurrent expenditures. In effect,this is a very strict interpretation of Hartwick’srule where all revenues from mining arereinvested. While spending under government’scapital budget does not ensure that allinvestment is productive, the SBI represents onetype of indicator, based on information that can

be provided by the SEEA, that can help tomonitor sustainability.

3.2.C Other socio-economic objectives

Countries may choose to sacrifice economicefficiency in order to achieve other importantsocio-economic objectives. Norway has chosento support small-scale fisheries as a componentof its strategy to promote regional development.Fisheries are a mechanism to create employmentand generate income in parts of the country thathave few options for employment. Norway iswilling to sacrifice economic efficiency and thegreater income this would generate in order toachieve this goal (Sorenson and Hass 1998).

Resource rent may be used to achieve a moreequitable distribution of benefits from the use ofresources, especially for economies that relyheavily on extractive industries. Within thecurrent generation, the resource rent can beused to support economic development thatbetters the lives of all citizens, not only theminority who may own companies. To ensureinter-generational equity, countries need toresist the pressure to consume all the income. Atleast some portion of the rent must be re-invested to contribute to increased well-beingfor future generations, as discussed above.

An analysis of access to resources and thebeneficiaries of the resource rent (the portionthat does not accrue to government) is anotherimportant aspect of resource management(Manning 2000). The terms of access by foreignoperators to a country’s resources may haveimportant implications for domesticemployment and income. Monitoring thissituation requires estimating the share of rentthat accrues to domestic operators, togovernment through taxes, and to foreignoperators. Where there are joint venturesbetween domestic and foreign companies, itmay be exceedingly difficult to determine theseshares.



EVALUATING TRADE-OFFS AMONG ALTERNATIVE

POLICY OBJECTIVES

When the pursuit of socio-economic goalsconflicts with economic efficiency, there is acost, and policy is more effective when this costis known. The costs of a policy which distributesaccess to resources more widely in society, forexample, but results in less efficientexploitation, can be measured as the resourcerent that has been sacrificed, and thecorresponding, lower value of national wealth,i.e., as the difference between the potential rentand rent actually generated. An initial, staticanalysis of the trade-off might measure this lossof rent based simply on information from microdata sets of individual companies under anexisting resource management regime, such asthe study of Norway’s herring fishery. Moresophisticated modeling of an industry would berequired to determine the long-term economiceffects of alternative management strategies onthe value of a resource, and the benefitsobtained from achieving other goals. Economy-wide modeling would be necessary to take intoaccount all the changes that would result fromalternative resource management policies.

DEVELOPING A CONSISTENT POLICY FOR RESOURCE

MANAGEMENT

As the case of Norway shows, resourcemanagement may be motivated by differentobjectives and result in very different outcomesin terms of efficiency, sustainability, and equity.Countries may well use different resources toachieve a range of socio-economic objectives—some resources managed purely commerciallyand others not. However, in some instances,policies for managing resources may have beendetermined independently for each resourcewithout the benefit of an economy-wide reviewwhich would establish a comprehensive policyfor all resources. This may have occurred in thepast because there was no comparableframework for measuring and analyzing allresources. It is particularly important for certainresources, such as forests, which have multipleuses cutting across different economic activities.Thus, comparing the management of allresources in the common framework of theSEEA provides a valuable tool for more rationalresource management.


4Physical Flow Accountsfor Pollution and Material Use

In the national accounts, the flow accounts areused much more extensively than the assetaccounts for economic analysis, and the samehas been true of the SEEA flow accounts forpollution and materials (including energy). Theflow accounts provide indicators ofsustainability as well as more detailedinformation to support economic analysis ofsources of environmental pressure and optionsfor change that can be used to improve theaggregate indicators of sustainability.

The SEEA flow accounts have two components:the physical accounts and the monetaryaccounts, which are constructed by valuation ofthe physical accounts. The physical accountshelp set priorities for policy based on the volumeof pollution, while the monetary accountsidentify the relative costs and benefits of reducingpollution. The flow accounts are also used ineconomic models to evaluate options fordealing with the problems: long-term strategiesfor addressing environmental problems and thepolicies for implementing a given strategy, suchas green taxes. This section describesapplications for deriving indicators as well asfor analysis first of the physical flow accounts,then of the monetary accounts.

4.1 Physical flow accounts

4.1.A Indicators and descriptive statistics

Data from the physical flow accounts are usedto assess pressure on the environment and toevaluate alternative options for reducingpressure on the environment. At their most

simple, the flow accounts monitor the timetrend of resource use, pollution emissions, andenvironmental degradation, both total and byindustry. A rising level of emissions, forexample, would be a clear warning sign ofenvironmental problems.

The overview of environmental trends helpsassess whether national goals, typically set interms of total figures for emissions or materialuse, are being achieved. A great deal of workhas been done throughout the industrializedworld to construct time series of pollutionemissions and energy use. Similar work hasbeen done for water accounts by a number ofcountries, including Chile, France, Spain,Moldova, South Africa, Botswana, Namibia, andthe Philippines. The example for Botswanashows declining per capita water use, anddeclining water intensity of the economy(measured by the GDP per cubic meter of waterused); the volume of water has still increasedbecause population and GDP growth outweighthe gains in efficiency (Table 6).

The aggregate indicators provide an overviewof the relationship between economicdevelopment and the environment; the moredetailed accounts help explain the overview.The construction and analysis of detailedstatistics are required to determine the roles ofdifferent factors in causing environmentalproblems and to design effective strategies toaddress them. Levels of pollution, material use,and waste are determined not only by the sizeof an economy, but by other factors as well,



including economic structure and technology.The more extensive the range of environmentalproblems and the sources of pollution, the moreimportant it is to develop detailed statistics.

The construction of environmental-economicprofiles, or “eco-efficiency” indicators hasbecome a common way of monitoring andranking industries in terms of theirenvironmental performance. There are manyversions of environmental-economic profiles,each showing some aspect of the environmentalburden imposed by a sector relative to theeconomic contribution it makes. Mostcommonly, the environmental burden isrepresented by a sector’s share of pollutiongenerated and economic contribution by itsshare of value-added, employment, andsometimes export earnings.

In the Netherlands, the profile revealed a veryunequal sectoral distribution of economicbenefits and environmental burdens (See Table7). The economic contribution for economicactivities is represented by value-added (GDP atfactor cost), and for final consumers, as totalconsumption expenditure. Reading across thefirst few rows of Table 5, it is clear thatindustrial activity accounts for most pollution(73–97%). Among industries, a large share ofpollution is produced by a relatively fewindustries, not at all in proportion to theireconomic contribution. For example, threeindustries account for 55% of greenhouse gasemissions: agriculture (15%), chemicals (14%),and public utilities (26%), but their combined

contribution to GDP is only 7%. Agriculturealone accounts for 47% of acidification and 91%of eutrophication emissions, but only 3% ofGDP. The striking environmental burdenimposed by agriculture compared to itsrelatively low economic contribution made theheadlines in Dutch newspapers and elsewhere.

These profiles have been constructed by allindustrialized countries with environmentalflow accounts (many were presented in a specialissue of SCED 1999), who have a longer historyof environmental monitoring. Economic-economic profiles for air pollution are used inNorway for “benchmarking” industryperformance, both for national environmentalpolicy as well as environmental management atthe company level (Hass and Sorenson 1998 ).The environmental performance of companiesor industries is compared to cleaner, moreefficient companies or industries in the samecountry or in other countries. The performanceof an industry over time is also monitored.Developing countries, perhaps becausemonitoring and benchmarking are relativelynew, have been slower to introduce thisindicator, with the exception of Chile, Botswana,Namibia, and South Africa for water, and thePhilippines and Korea for pollution.

While the environmental economic profiledescribed above is useful for cross-sectoralcomparisons, it is cumbersome for comparisonof performance over time or across countries.For such comparisons, a different indicator, thepollution or material intensity of a sector, may

Source: Lange and others 2000.

Table 6. Index of water use, GDP growth and population growth in Botswana,

1993 to 1998 (1993 = 1.00)

1993/94 1994/95 1995/96 1996/97 1997/98 1998/99

Volume of water used 1.00 1.01 1.03 0.99 1.04 1.05

Per capita water use 1.00 0.99 0.98 0.92 0.94 0.93

GDP per m3 waterused 1.00 1.02 1.06 1.18 1.22 1.26


Physical Flow Accounts for Pollution and Material Use

be used. This is calculated as the quantity ofsectoral emissions per unit of sectoral output orvalue-added. Such indicators for pollution havebeen used widely within industrializedcountries to assess performance over time,although there have been no formal cross-country comparisons.

Table 8 shows the comparative economiccontribution of water use in three neighboring,

water-scarce countries: Botswana, Namibia andSouth Africa, which all share internationalwaters. While the striking differences cannot befully explained here, agricultural policy is amajor factor: Botswana has relatively littleagriculture, while Namibia and South Africahave, until recently, subsidized low value,irrigated agriculture. These profiles, on a moredisaggregated sectoral level, have been usedboth for internal policy analysis and in on-going

Table 7. Net contribution of consumption and production to GDP and to six

environmental themes in the Netherlands, 1993

Economy Environment

Greenhouseeffect

Ozone-layer

depletionAcidifica-

tionEutrophica-

tionSolidwaste

TOTAL, % 100 100 100 100 100

Consumption 19 2 15 9 3

Industry 79 97 85 91 66

Capital and other sources 2 1 - - 31

HOUSEHOLD CONSUMPTION,% of spending and residuals

100

100 100 100 100 100

Own transport 8 38 - 88 21 1

Other consumption 92 62 100 12 79 99

INDUSTRY, % of GDP andresiduals 100 100 100 100 100 100

Agriculture, hunting, forestry,

fishing 3 15 2 47 91 7

Mining and quarrying 3 2 - 1 - 1

Manufacturing

Petroleum industry 1 7 - 11 -

Chemical industry 2 14 27 6 2 16

Metal products and

machinery industry 3 2 9 1 - 2

Other manufacturing 12 12 20 7 6 25

Public utilities 2 26 - 9 1 2

Transport and storage 6 8 6 12 1 5

Other services 68 14 36 6 -1 42

Source: Statistics Netherlands (EPIS-report).



negotiations over regional allocations of sharedwater. Identifying the economic contribution ofwater in agriculture relative to other sectors hasprovided policymakers in southern Africa witha more sound basis on which to discuss futurepolicy for agriculture as well as other sectors.

The level of pollution from a given sectordepends on two factors: pollution intensity(determined by its technology) and level ofoutput. Reducing pollution can be achieved byaddressing either or both of these factors. Whilethe eco-efficiency profile shown above reportsthe direct generation of pollution associated withproduction, it is important for policy makers tounderstand the driving forces that cause suchlevels of pollution. In the Dutch example, arelatively small proportion of total residuals isgenerated by final consumers; most pollutionresults from industrial production. However,production takes place in order to supply otherindustries with inputs they need and,ultimately, to supply final users with productsthey want.

Input-output (IO) analysis is used to calculatethe total (direct + indirect) impact of deliveringa unit of a sector’s output to final users. Ineffect, this analysis redistributes emissions fromindustry to the driving force (final user) for

which this production took place.Long experience withenvironmental input-outputanalysis has shown that the totalimpact is often much larger thanthe direct impact (Førsund 1985,Miller and Blair 1985, Pearson1989).

Figure 7 provides a comparisonof the direct and total emissionsof sulfur dioxide for eachproduct delivered to final usersin Sweden. For every one millionkroner purchase of agricultural

output by final users 50 kg of SO2 are generateddirectly by the agricultural industry, and anadditional 50 kg of SO2 (for a total of 100 kg) isgenerated indirectly. Similar statistics have beenconstructed for other industrialized countrieswith environmental accounts, but, again, thisaspect of the flow accounts has beenunderutilized by most developing countries.

The ability to measure total pollution and use ofmaterials—direct plus indirect—associated withgiven products, industrial processes, orconsumption patterns makes possible far moreeffective strategies for reducing the use, andmanaging the disposal, of materials than couldbe devised on the basis of direct use only. Forexample, Table 7 showed that public utilities,mainly electricity production, was responsiblefor 24% of greenhouse gas emissions and 9% ofacidification emissions in the Netherlands. Inattempting to reduce these emissions, policyincentives can be designed to bring abouttechnological change in the electric powerindustry to reduce its direct emission coefficient,but policy can also attempt to change thebehavior of those who purchase electricity. It isoften necessary to design policy for bothgroups—the direct source of pollution as well asthe users of products, whose demand drives thelevel of production and associated pollution.

Table 8. National income per cubic meter of water by sector in

Botswana, Namibia, and South Africa, 1996

(Botswana pula per cubic meter of water used)

Note: Calculated as water input divided by sectoral value-added.

Source: Lange and others 2000.

Botswana Namibia South Africa

Agriculture 9 6 2

Mining 420 54 44

Manufacturing 437 189 98

Trade, Services, Government 724 542 302

GDP per m3 of water input 124 45 20



The calculation of total emissions is particularlyuseful for understanding how the structure ofan economy affects the levels of pollutionemissions and resource use. Final use can bedisaggregated into its components—householdconsumption, government consumption,investment, exports and imports—to determinehow much of the total pollution generated inthe economy occurs in order to meet thedemands of each of these components. Becausehousehold consumption accounts for thegreatest share of final demand, researchers haveincreasingly focused on the composition ofhousehold consumption as a critical componentfor sustainable development. Strategies forsustainable development have examined theimpact of alternative household consumptionpatterns, particularly in the wealthyindustrialized countries where the need to

identify “sustainable lifestyles” has received alot of attention.

Over time, levels of emissions can changeconsiderably, and policy makers need to knowhow much of this might be the result ofenvironmental policies or other factors. Policiesaffect the choice of technology (direct andindirect emission intensities) and the level andcomposition of final demand (driving forces),and it is not immediately evident how much ofthe change in emissions is attributable to eachfactor. Structural decomposition analysis is aformal technique developed to distinguish thedifferent sources of change in the economy overtime, including a) effects originating fromchanges in the structure of final demand versuschanges in intermediate input coefficients, andb) further decomposition to distinguish between

0 50 100 150 200 250 300 350 400 450

Transportation

Electricity

Iron & steel

Chemicals

Pulp & paper

Food, textiles, wood

Mining

Fishing

Agriculture

kg of SO per million SEK of industry output2

Direct + Indirect

Direct

Figure 7. Direct and total emissions of sulfur dioxide per unit of

industrial output delivered to final users in Sweden, 1991

Source: Hellsten and others 1999.



effects of relative prices (substitution) andtechnological change. This analysis is carriedout by decomposing differences in the total(direct plus indirect) requirements matricesderived from the IO tables.

Decomposition analysis using the NAMEA forthe Netherlands, was carried out for the period1987 to 1998 to address changes in the levels ofgreenhouse gas emissions, acidificationemissions, and solid waste. The results for CO2show that economic growth (the volume ofproduction, +35%) outweighed the impact ofimproved efficiency (-11.5) and structuralchange (the changing composition of finaldemand, -3.2%) for greenhouse gas emissions,resulting in a 20.3% increase in CO2 emissionsover the period (Table 9).

4.1.B Policy analysis and strategicplanning

So far, the discussion has focused on analysis ofthe interaction between the existingenvironment and economy, but policy-makersalso need to anticipate the future and designeffective instruments for environmental policy.These concerns are typically addressed in amulti-sectoral, economy-wide modelingframework. The use of the SEEA flow accountsfor environmental-economic modeling isdiscussed here. Some modeling is undertakenprimarily to derive environmentally-adjustedmacroeconomic indicators, such as Hueting’sSNI. These modeling applications are discussedin section 6 along with other macroeconomicindicators.

The process of analysis has two majorcomponents: strategic analysis in order todesign a more desirable future and policyanalysis in order to implement such strategies.Countries typically identify the broadenvironmental objectives they wish to achievein the future, such as more sustainabledevelopment or more sustainable lifestyles.Under these objectives, specific problems, suchas air quality, would be identified and moredetailed strategies to address these problems areexamined in a modeling framework. Thisanalysis is often based on long-term models toexplore alternative scenarios about paths ofeconomic development, including quite bold,speculative alternatives.

Once the overall strategies are determined, thebest policy measures to implement thestrategies need to be identified. This analysisconsiders different instruments policy makersmight take (usually a modest range of actionsover a relatively short period of time, such asdifferent values for a carbon tax), and providespolicy makers with the most likely outcome ofthese actions. Policy models can be used toexamine the economic implications of variousenvironmental policy instruments—such astaxes, tradable permits, or emission standards—as well as macroeconomic policies, such as tradepolicy and its impact on the environment.

STRATEGIC ANALYSIS

Strategic planning addresses alternativedevelopment paths over a relatively long-termtime horizon (10–25 years or more) and thefundamental changes in the structure of theeconomy that might be necessary to achievesociety’s environmental objectives. Thee modelsare used to project levels of emissions, ormaterial requirements and the long-termeconomic impact of environmental policies. Itconsiders new technologies that might beintroduced over a long period and the changesin final demand, especially private

Table 9. Decomposition of the percent change in

CO2 between 1987 and 1998, the Netherlands

(percentage change from 1987 to 1998)

Source: De Haan (undated paper).

Volume Structure Efficiency Total

35.0 -3.2 -11.5 20.3



consumption. Strategic planning oftenemphasizes dynamic modeling instead of thestatic analysis commonly used for policyanalysis. Dynamic analysis is important becauseit informs policymakers about the transitionpath—the process of adjustment—to a neweconomy.

Examples of strategic planning include theNetherlands Environmental Policy Plan, a long-term plan for sustainable development.Norway’s Ministry of Finance has long used astrategic macroeconomic planning model, theMulti-Sector Growth Model, which hasintegrated environmental components (energyand pollution accounts) from the SEEA. A greatdeal of work has been done with global modelsto project different global trajectories forgreenhouse gas emissions. A few one-timestudies have been carried out usingenvironmental accounts in developing countriesbut are not yet fully integrated in their planningprocesses. For example, a multi-sector, dynamicIO model was used to assess the environmentalimplications of Indonesia’s Second Long TermDevelopment Plan for the Ministry of Planning(Lange 1997).

POLICY ANALYSIS

The flow accounts are widely used for policyanalysis, for example, to assess the impact ofenvironmental tax reform, to design economicinstruments to reduce pollution emissions, andto assess competitiveness under new, morerestrictive environmental policies. The EU hasbeen the largest user of the accounts and hasused them mainly to address two priorities:greenhouse gas emissions and acid rain.

Norway has used the flow accounts for energyand greenhouse gas emissions to assess a policythat many countries are considering: changingthe structure of taxes to increase taxes onemissions and/or resource use, whilesimultaneously reducing other taxes by an equalamount in order to remain fiscally neutral, the

so-called “double dividend” (Bye 1997,summarized in Natural Resources and theEnvironment 1998). Norway used its multi-sector general equilibrium model to lookspecifically at increasing the carbon tax to NOK700 per ton of CO2 with a compensatingdecrease in its payroll tax. Policymakers inNorway wanted to know what effects this taxreform would have on economic welfare.

Using a multi-sector, general equilibrium modelof the economy, Norway initially found thatemployment and economic welfare wouldincrease while carbon emissions declined.However, closer analysis of the results indicatedthat the tax reform would result in significantstructural change in the economy—certainenergy-intensive industries in the metal,chemical and oil refining industries wereparticularly hard hit by the tax, and wouldreduce output and employment considerably.Furthermore, these industries weredisproportionately located in small towns wherean industry might be the only major employer. Itis reasonable to assume that, at least in the shortterm, people would be reluctant to move to newtowns in search of new jobs. By including thiselement of labor immobility, the model showedthat although emissions would still decline, theeconomic improvement disappeared andeconomic welfare actually declined slightly.

In considering environmental taxes, anotherissue policymakers must consider is the impactsuch taxes might have on the internationalcompetitiveness of their domestic industries.This is an especially important issue for veryopen economies like the Netherlands. A studyaddressing this issue for the Netherlands wasundertaken using the flow accounts for energyas well as for carbon emissions, othergreenhouse gas emissions, acidification andeutrophication emissions (Komen and Peerlings1999). The study quantified the relativesensitivity of the different industries to changesin environmental taxes.



In another example, Swedish researchers wereable to show policymakers that policies toreduce carbon emissions may generateadditional unintended (or ancillary) benefitsthat should be taken into account whenconsidering the advantages and disadvantagesof different environmental policy reforms(Nilsson and Huhtala 2000). The study analyzedthe advantages of utilizing a system of carbontrading permits in order to meet Sweden’scarbon target under the Kyoto Protocol as analternative to implementing measures to reducedomestic levels of carbon emissions. When onlythe benefits of reduced carbon emissions wereconsidered, the purchase of low-cost carbonemission permits was the more cost-effectivemeans of meeting Sweden’s targets. However,measures to reduce domestic emissions ofcarbon also resulted in lower emissions of sulfurand nitrogen at no extra cost. When thisancillary benefit was taken into account, thepurchase of carbon emission permits was not asadvantageous as measures to reduce domesticcarbon emissions. Many other countries haveundertaken similar studies to the threedescribed here.

In other countries, environmental priorities maybe quite different, but similar modelingtechniques can be applied. For example, in partsof Australia, the United States, and southernAfrica, water scarcity is as critical an issue aswater quality. The flow accounts for water wereused in a CGE model to address new waterpricing policies in South Africa (Hassan 1998).In the past, water prices were very low, withlittle regard for cost or scarcity, especially foragricultural use. The proposed new pricingpolicy includes full-cost recovery tariffs with aguaranteed “lifeline” amount of water suppliedto all households, as well as innovative pricingpolicies such as a user charge for the reductionof rainfall runoff caused by commercialplantations of exotic forest species.

A broad range of policy studies was undertakenin the Philippines, including an assessment ofthe environmental implications of rice self-sufficiency (land, water, and pollution), a studyof trade and environment linkages, and a studyof the environmental implications of alternativeland use patterns (ENRAP 1998).

Other applications of the physical flow accountsinclude life-cycle analysis (LCA). Traditionally,LCA is a bottom-up process analysis, based onlinking the specific processes in a supply chainto trace the environmental impacts from “cradleto grave” of specific products, or productionprocesses. The advantage of this extremelydetailed approach is its capacity to representenvironmental impacts precisely. However, amajor limitation of process-based LCA is thelikelihood that important parts of the productsystems are left out of the analysis, simplybecause it is a very difficult task to follow theentire supply chain in such detail.

In some instances, practitioners have attemptedto address this problem through the use of so-called hybrid life-cycle analysis in which thedetailed, partial LCA is combined witheconomy-wide IO-analysis. Physical flowaccounts for inputs of natural resources andoutputs of residuals are combined with input-output analysis as a supplement to traditionalprocess-oriented LCA.

A recent workshop sponsored by the DanishMinistry of Environment and Energy presentedstudies from a number of countries (Nielsen andWeidema 2001). This method was used, forexample, to calculate the total CO2 impact fromDanish household consumption of 72 differentcommodities (Munksgaard 2001). This methodhas also been applied to examine theenvironmental implications of introducing fuelcell electric vehicles in the United States (Gloria2001). The combination of traditional LCA withIO has been used in the UK to explore aspects of



climate change policies, such as thedevelopment of a model for carbon emissionstrading and an analysis to identify theindustries that would gain and lose the mostfrom carbon taxes (Ecotec 1999).

MATERIAL FLOW ACCOUNTS

MFA were developed by the Wuppertal Institute(Spangenberg and others 1999) to addresssustainability based on the ecological concept ofde-materialization, or de-linking economicgrowth from material use. MFA are similar tothe SEEA’s physical flow accounts in that theyrecord the use of materials and the generation ofresiduals, but MFA are not always fullydisaggregated by industry. The SEEA canimprove MFA by providing information aboutmuch of the material use disaggregated byindustry. However, MFA also include the“hidden flows,” which consist of materialsexcavated or disturbed along with the desiredmaterial, but which do not themselves enter theeconomy. Examples of hidden flows includemine tailings or soil excavated duringconstruction. MFA also distinguish betweendissipative use flows and flows which areembodied in products. Dissipative flows arematerials which are shed from products duringthe normal course of use, such as fertilizer, orrubber worn from motor vehicle tires.

MFA have been constructed by a number ofcountries. An ambitious project by the WorldResources Institute has made a substantialimprovement by compiling roughly comparableMFA for five industrialized countries: Austria,Germany, Japan, the Netherlands, and theUnited States (Matthew and others 2000). Inaddition to compiling macro-level indicators,the study analyzed the flows to investigateissues such as the impact of e-commerce and theshift from heavy industry in the five countriesto knowledge-based and service industries. Thestudy found that despite these changes and theincreasing efficiency of material use, the use of

materials and the volume of waste generatedcontinues to increase. The authors attribute thisto economic growth and consumer choices thatfavor energy- and material-intensive lifestyles.

EXTENSION FROM NATIONAL TO

REGIONAL ANALYSIS

A country trying to design a more sustainableeconomy with its own environmental accountscan face two problems. First, even if it reducesdomestic emissions, it may still sufferenvironmental damage because oftransboundary flows of emissions from othercountries. This is the case for regional issuessuch as acidification of air and eutrophication ofwater, as well as global issues such as climatechange and ozone depletion. Construction ofnational accounts may need to include theinternational transfers of residuals. This hasbeen proposed in the SEEA, but will not be easyto implement. Europe has a good system formonitoring transboundary flows, but manyother parts of the world do not.

Secondly, world trade has led to a dissociationof consumption patterns that causeenvironmental degradation from the source ofproduction where the degradation occurs. Openeconomies typically import a great deal, and thepollutants associated with the production ofthese imports in their countries of origin mightbe quite high. Attempts to take imports intoaccount when assessing a country’s totalenvironmental impact have been hampered bylack of information about the emissioncharacteristics of trading partners. Some studieshave simply assumed that the imports have thesame pollution and energy coefficients per unitof output as domestically produced products.The construction of flow accounts by manycountries will make it possible to substitute realinformation for this highly unlikely assumption.This would greatly improve the estimates of thetrue environmental burden of a country’sconsumption patterns.



To estimate the importance of using country-specific emission characteristics, Swedencompared the emissions embodied in theirimports under three alternative assumptions: 1)assuming Swedish industry-level emissioncoefficients for imports; 2) assuming nationalaverage emission intensity for all imports fromeach country based on emissions data for all EUcountries plus other major trading partners(USA, Japan, Norway, and Switzerland); and 3)assuming industry-specific emissionscoefficients for each country, derived from theirenvironmental accounts.

The results for CO2, SO2, and NOx emissionsembodied in imports were extremely sensitiveto the method used (Table 10)1. The lowestestimate of CO2 and SO2 emissions wasobtained when Swedish emission coefficientswere used: emissions using the other twomethods were at least 50% higher. The reverseoccurred for NOx: Swedish emission coefficientsresulted in the highest level of NOx emissions,although not much higher than the other twomethods. The results reveal significantdifferences among countries in emissionintensities. While there are a number ofmethodological and data improvements needed,this pilot study indicates the importance ofobtaining environmental accounts for all majortrading partners in order to evaluate correctlythe emissions embodied in imports.

4.2 Monetary flow accounts forenvironmental degradation and resourceuse

Effective environmental management is basednot only on an understanding of the volume ofpollution and material use, but also anunderstanding of the economic implications.Policy makers need to know what the welfareloss of pollution is (damage costs) and wherelimited financial resources will be most effectivein reducing environmental pressure, i.e., therelative benefits and costs of reducing differentforms of environmental degradation fromdifferent sources.

Two different conceptual approaches to valuingenvironmental degradation are commonly used:the maintenance cost approach and the damagecost approach. The former shows policymakersthe cost of certain actions to prevent orremediate degradation, and the latter shows thebenefit of policy actions, that is, the value of thedamages that will be prevented. In the absenceof efficient markets, these measures are likely tobe quite different. The damage cost is thetheoretically correct approach for measuringchanges in economic well-being and adjustingmacroeconomic aggregates, but both measuresprovide useful information for environmentalmanagement.

4.2.A Indicators anddescriptive statistics

Many of the indicators anddescriptive statisticsdeveloped for the physicalaccounts can be used for themonetary accounts. Trends inaggregate monetary figurescan be monitored as well asmeasures of pollutionintensity by economicactivity. Variousenvironmental-economic

Table 10. Emissions embodied in Swedish imports under alternative

assumptions about emission intensities of imports, 1995

Note: Method 3 used data from 1993 and so is not directly comparable to the other two

methods.

Source: NIER 2000a.

CO2 SO2 NOx

Method 1: Swedish emission coefficients

20,800 43 128

Method 2: National average emission coefficients

from exporting country 32,900 121 119

Method 3: Industry-specific emission coefficients

from exporting country 36,300 128 109



profiles can be constructed, and the monetaryaccounts can be analyzed to determine directand indirect sectoral emissions, and how thesemay have changed over time. An advantage ofthe monetary accounts is that theenvironmental-economic profiles can aggregatedifferent kinds of environmental problems intoa single figure to represent the totalenvironmental burden from each sector.

Figure 8 provides a snapshot for Sweden of theoverall economic contribution andenvironmental burden posed by the mostpolluting industries in 1991. The environmentalburden is represented by sectoral shares ofdamages caused by domestic emissions ofsulfur dioxide (SO2), nitrogen oxides (NOX),volatile organic compounds (VOC), nitrogen towater, and ammonia (NH3). The economiccontribution is represented by sectoral shares ofGDP and employment. The highest burden is

imposed by Transportation (33%), followed byServices (9%), Pulp and paper (7%), andAgriculture (7%). However, the Serviceindustry’s economic contribution (36% of GDPand 43% of employment is) much higher thanits share of the environmental burden. Thereverse is true for Transportation whoseeconomic contribution is only 6% of nationalincome and 9%.

Information about relative economiccontributions and environmental burdens isessential for policymakers when identifyingindustries that will play a key role in economicdevelopment. In the absence of suchinformation, incentives to promote growth of aspecific industry, such as subsidies to Pulp andpaper or Agriculture, may result in levels ofenvironmental damage that far outweighapparent economic gains.

0

5

10

15

20

25

30

35

40

45

Pe

rce

nt

Agric

ulture

Food, text

ile, s t

one, wood

prod.

Pulp

, paper

Chem

ica l

Eng in

eering

Ele

c., ga s ,

hea ting

Cons t

ruct

ion

Transp

ortatio

n

Serv

ices

Damage cost

GDP

Employment

Source: Ahlroth 2000c.

Figure 8. Economic contribution and environmental burden from

domestic pollution by selected industries in Sweden, 1991



In cases where environmental policy takes theform of setting emission standards withoutregard for balancing economic costs andbenefits, the policy challenge is to find the mostcost-effective opportunities for pollutionreduction, represented by the maintenance costapproach. This approach was used in thePhilippines and in Korea.

Korea was one of the first developing countriesto construct environmental accounts and washampered by a lack of data. Monetary accountswere constructed assuming the same pollutionabatement method and unit cost for allindustries, except for a few adjustments formobile sources of air pollution (KoreaEnvironment Institute 1998). Consequently, thepattern of damage costs by industry roughlyparallels the pattern of physical emissions andthe accounts does not provide policy-makerswith any additional information forenvironmental management.

The Philippines constructed environmentaldegradation accounts for a range of pollutants

to air and water from selected industries, aswell as nutrient loss in agriculture and soil lossin forestry (NSCB 1998, 2000). Some resultsfrom the accounts for biological oxygen demand(BOD) are shown in Table 11. The results areinstructive: although aquaculture is responsiblefor 64% of total BOD emissions, the cost ofpollution abatement in that industry isextremely small—less than 1% of total costs. Bycontrast, the hog industry produces 34% of BODbut accounts for nearly 80% of the maintenancecosts. The sugar industry contributes a tinyshare of total BOD emissions, only 0.4%, but itwould be quite costly to reduce these emissions:its share of environmental damage costs is 13%.

Comparison of valuation approachesTo compare the differences among valuationtechniques, Sweden applied three techniques tothe NOX: the damage cost approach, theavoidance cost approach (another term formaintenance cost), and willingness-to-pay(Table 12). The damage cost approach yields thelowest value, at just over 2,000 million SEK; theavoidance cost approach is higher, but not that

dissimilar, at around 2,800million SEK. At 7,300 millionSEK, WTP yields a value morethan three times the damagecost estimate.

While estimates of bothdamage and maintenance costare available for someEuropean countries (seediscussion of GARP projectbelow), few developingcountries have implementedthe damage cost approach,except in some cases for thePhilippines (see ENRAP(1998)). The primary reasonseems to be data availability.To estimate maintenance cost,many countries can use thebenefits transfer method, i.e.,

Table 11. Emissions of BOD and environmental damage by selected

industries in the Philippines, 1993

Na: not applicable

Notes: Environmental damage estimated using the maintenance cost approach.

Emissions of BOD were not calculated for all industries.

Source: National Statistical Coordination Board 2000.

Percent ofemissions

Percent ofenvironmental

damage

Ratio of costshares to

emission shares

Aquaculture 63.7 0.7 0.01

Hog industry 34.2 79.7 2.33

Tuna canning 0.1 0.3 2.86

Textile industry 1.4 5.8 4.02

Leather tanning 0.1 0.3 5.03

Sugar industry 0.4 13.2 31.09

Total 100.0 100.0 na

Level of emissions (MT) and

total costs (thousands of pesos) 1,303,452 2,053,000



using estimates developed by other countries. Alarge databank for the cost of different pollutionabatement technologies, which can be appliedin virtually any country, is readily available.Environmental damage, however, is lessamenable to benefits transfer because it is oftena country-specific matter, both in terms of therelationship between level of emissions andamount of damage, and the local cost of thisdamage. Until more easily usable, orinternationally transferable estimates of valueare available, it will be difficult for manydeveloping countries to implement the damagecost approach.

TRANSBOUNDARY POLLUTION

Ideally, the two valuation methods wouldprovide a crude indication of the relativebenefits and costs of reducing emissions (thislimitations to thiscomparison are discussedin the revised SEEA).However, there is not asimple correspondencebetween domesticdamage and domesticmaintenance costs, inpart, because of the largerole played bytransboundary pollution.Like many countries,Sweden imports andexports a great deal ofpollution—more than60% of its domesticproduction of NOX isexported, and imported

NOX accounts for 72% of its net domesticdeposition (Figure 9).

The very high share of imported emissionsindicates that Sweden will need to cooperatewith neighboring countries, who are sources ofits imported NOX, in order to improveSweden’s domestic environment. The crudeevidence from these accounts of the relativecosts and benefits of reducing domesticemissions in Sweden lends further support tothat approach. However, given Sweden’sconsiderable exports of NOX, an efficientstrategy for reducing emissions must take intoaccount all countries involved, in order toidentify in a regional context where the greatestbenefits at lowest cost are to be found.

An ambitious study to map the costs oftransboundary pollution among 15 EU countrieswas undertaken by the GARP project. Table 13shows the distribution of these costs: thedamage due to domestically generatedresiduals, the damage due to imports, and thedamage that is exported. The first row, labeledEU, shows a total damage cost of 128.8 billioneuros per year. Germany, France, UK, and Italyare the source of most of this pollution, and

Table 12. Cost of NOx

emissions

using different valuation methods

in Sweden, 1991 (millions of SEK)


Damagecost

Avoidancecost

Willingness-to-pay

2,020 2,792 7,301

Figure 9. Domestic emissions, exports and imports of NOX in

Sweden, 1991


0

100

200

300

400

500

600

Domestic sources Exports Imports Net domesticdepostion

ktons

ofN

OX



these four countries also bear most of the cost.For many countries, domestic emissions areresponsible for less than half the total domesticdamages. Smaller countries with manyneighbors tend to incur a high share of damagesfrom imports, while larger countries sufferdamages mainly from domestic residuals.

4.2.B Policy analysis with monetaryaccounts

As with the physical flow accounts, themonetary accounts can have many applications.They were used in the Philippines for partialequilibrium cost-benefit analyses of the use ofeconomic instruments for addressing airpollution. One cost-benefit analysis consideredtwo alternative policies to reduce atmosphericlead: a tax on leaded gasoline vs. a completephase-out of leaded gasoline over a three-yearperiod (Manasan et al. 1998). The accountsprovided the physical lead emissions, dataabout the value of benefits (improved human

health due to lower emissions) and costs(measures to reduce emissions) were obtainedfrom other studies and used to constructmonetary accounts. The results found that thepresent value of the phase-out was close to threetimes that of the tax differential approach,primarily because of its much greater and fasterimpact on health. For this reason, the authorsconclude that the phase-out is a moreappropriate policy than using economicincentives to reduce lead emissions.

The monetary accounts could potentially beused for more comprehensive policy analysis.For example, the abatement costs could providethe input into models such as the one used tocalculate Hueting’s Sustainable NationalIncome, or models evaluating green taxes wherethe measures industries take in response totaxes would be included. In practice, the datarequired by these models is not obtained frommonetary flow accounts because entire

Table 13. Pollution damage by source country and receptor country in the European Union

Source: Markandya and others 2000.

Receptor Countries

AT BE DE DK ES FI FR GR IE IT LU NL PT SE UK EU NonEU

Damage Costs caused by the Source Countries within the Receptor Countries [billion ECU/a]

EU 2.8 4.5 40.9 2.3 8.9 0.4 21.4 2.0 0.4 15.3 0.1 7.0 1.2 2.1 19.4 128.8 34.9

Percentage of Damage Costs Caused in the Receptor Countries [%] [bn ECU/a]

AT 12.2 0.2 0.9 0.6 0.2 0.9 0.2 0.9 0.1 2.2 0.3 0.2 0.0 0.8 0.1 1.2 1.8

BE 1.1 12.3 3.7 3.8 1.2 1.4 3.7 0.0 1.4 0.7 4.8 13.0 0.4 2.6 1.1 4.4 0.4

DE 47.0 14.2 53.8 43.7 4.8 38.7 13.4 2.0 6.9 15.6 33.3 15.6 0.9 49.6 6.7 34.4 17.0

DK 0.6 1.0 0.9 9.2 0.1 4.6 0.5 0.0 0.6 0.1 0.8 1.0 0.0 8.9 0.8 1.2 0.4

ES 1.7 8.6 3.8 1.8 51.8 0.0 16.3 0.0 16.1 5.6 8.9 6.1 50.4 1.0 7.3 13.5 0.4

FI 0.1 0.1 0.1 0.4 0.0 29.5 0.0 0.0 0.1 0.0 0.0 0.1 0.0 1.9 0.1 0.3 0.1

FR 8.7 33.5 15.3 7.2 16.8 2.1 36.2 0.1 11.2 10.3 31.9 23.8 5.9 5.4 11.4 23.2 2.0

GR 1.1 0.0 0.1 0.2 0.1 0.0 0.1 78.0 0.1 2.3 0.1 0.0 0.0 0.2 0.0 2.1 3.7

IE 0.0 0.3 0.2 0.5 0.4 0.1 0.4 0.0 18.8 0.0 0.2 0.4 0.3 0.2 2.1 0.7 0.0

IT 23.3 3.0 6.9 2.5 8.8 1.6 5.9 18.9 2.8 60.1 4.1 2.4 2.6 2.8 1.2 15.8 6.8

LU 0.2 0.3 0.4 0.2 0.1 0.1 0.2 0.0 0.1 0.1 1.6 0.4 0.0 0.1 0.0 0.3 0.0

NL 1.2 8.0 4.6 7.1 1.0 2.3 3.8 0.0 1.4 0.6 5.6 13.8 0.3 5.0 1.8 4.9 0.5

PT 0.0 0.6 0.2 0.1 6.8 0.0 1.1 0.0 2.4 0.1 0.6 0.4 36.0 0.0 0.7 1.6 0.0

SE 0.2 0.3 0.4 1.5 0.0 11.7 0.1 0.0 0.2 0.0 0.2 0.3 0.0 8.1 0.3 0.5 0.3

UK 2.6 17.5 8.6 21.2 8.0 7.0 18.1 0.0 37.8 2.0 7.7 22.4 3.3 13.4 66.5 24.7 1.2



abatement curves need to be estimated. Themonetary accounts provide only a summaryfigure, although if the summary figures arederived from abatement cost curves, thisunderlying data may be useful in policy models.

Valuation issues discussed in environmentalaccounts, and in the SEEA chapter on valuation,have largely focused on environmentaldegradation, but policymakers can use thisapproach to address another challenge: pricingnon-market goods and environmental services.Two examples of particular importance fordeveloping countries have been water andrecreational services from nature-basedprotected areas. In many countries, the pricecharged for these resources does not reflect thetrue financial cost, let alone the full economiccost. Where the costs are subsidized, as is wateris in many countries, there is little incentive forresource conservation.

When tourists visit magnificent national parksand protected areas, they may not be charged anentrance or user fee that equals the full value ofthe benefits they receive. As a result, therecreational services and the unique ecosystemson which they depend, are undervalued. Giventhe apparently low economic value for this formof land use, countries often face pressure toconvert protected areas to other uses. Thisproblem can be especially severe for developingcountries, e.g., the clearing of protected areaforest for agriculture in many tropical countries,but is by no means limited to developingcountries, e.g., the controversy over theconversion of Alaska’s Arctic wilderness to oilmining.

There is an extensive literature on the economicvalue of both protected areas and water, butrelatively little has found its way into theenvironmental accounts yet. Where water rightsare traded in reasonably competitive markets,

as in parts of Australia, the value of water isreflected in the price of these rights. However,water is often not traded in competitive marketsand its value can be difficult to measure,requiring a great deal of information that is notalways readily available. Case studies of thevalue of water for agriculture were carried outin the environmental accounting framework forNamibia (Lange and others 2000), and found avery low value for water, though one thatvaried enormously by crop.

Even without estimating the economic value ofwater, there is monetary information about costsand tariffs that is very useful to policymakers.Flow accounts can be compiled for the cost ofproviding water to each sector, the tariffscharged, and, from this information, the subsidyby sector can be calculated. Monitoringsubsidies is clearly important both forsustainable management of resources as well asfor equity by identifying which groups insociety receive the greatest subsidy. This workhas been done, at least partly, in Chile, France,Botswana, and Namibia.

These calculations are not only important fordomestic environmental policy, but also forissues that are regional or global. For example,Namibia’s water subsidies to commercialagriculture have declined considerably since theearly 1990s when a policy of graduallyintroducing full-cost recovery was adopted. Acomparison of water subsidies for Namibia andSouth Africa in 1996 (Lange and Hassan 1999)showed that commercial agriculture in Namibiacontinued to receive significant water subsidies,though much lower than the subsidies tocommercial agriculture in South Africa.Quantifying the value of the subsidy has beenparticularly useful both in domestic discussionsabout water and agricultural policy, and also inthe negotiations over future allocation of sharedriver water between Namibia and South Africa.


Environmental Protection andResource Management Accounts5

Environmental regulation has been highlycontroversial in most countries and theenvironmental protection and resourcemanagement accounts were designed to helpaddress some of the important questionssurrounding regulation. For example, has themoney spent on pollution abatement beeneffective in reducing pollution, and how hasenvironmental regulation affected productivityand international competitiveness? This set ofaccounts has several quite distinct componentsincluding:1. Expenditures for environmental protection

and resource management, by public andprivate sectors

2. The activities of industries that provideenvironmental protection services

3. Environmental and resource taxes/subsidies.

Section 2 discussed a range of problems inconstructing and interpreting the EPEcomponent of the accounts that make it difficultto use them for policy. For example, adecreasing EPE cannot be interpretedunambiguously as a trend toward a more or lesssustainable economy. With more information toclarify some of the ambiguities if the EPE,practitioners can analyze some of the followingissues:• The cost of environmental regulation over

time• How effective environmental protection

expenditures and eco-taxes have been inreducing pollution

• The economic impact of environmentalexpenditures and taxes, for example, theimpact of carbon taxes on prices,productivity and internationalcompetitiveness.

The first part of this section addresses theconstruction of indicators and descriptivestatistics from these accounts to use formonitoring environmental protection andresource management activities. The secondsection discusses the use of the accounts foranalysis and policy modelling.

5.1 Indicators and descriptive statistics

Of the three components of this part of theaccounts, environmental protection expenditure(EPE) accounts have been the most widelyconstructed, mainly in the United States,Canada, the EU, Japan, and Australia. Somedeveloping countries have also constructed EPEaccounts, notably the Philippines, Korea,Colombia, and Chile.

Environmental protection expenditureaccounts

Eurostat has published a handbook with adetailed list of indicators that can be obtainedfrom the EPE accounts, from the most general(e.g., time trend of EPE by sector and domain)to detailed (e.g., spending within industries bydomain). EPE accounts for the United States, forexample, show that, as a percentage of GDP,expenditures have remained constant atbetween 1.7 and 1.8 percent. Spending by



domain (air, water, solid waste) was dominatedby spending for water pollution from 1972 to1994, but that spending for solid wastetreatment was growing fast. Most of thespending for pollution abatement and control isundertaken by the private sector. A largeamount of expenditures is undertaken bygovernment for water treatment and solid wastemanagement.

Of the four developing countries that havecompiled EPE, coverage differs from country tocountry. Only Colombia and the Korea coverEPE by all sectors. Costa Rica and thePhilippines have compiled only EPE bygovernment. Figures for Chile have not yet beenpublished. The indicators compiled from theiraccounts are shown in Table 14. It is not clearthat anything has been done with these figures.

A more detailed breakdown of expenditures byindustry can be used to identify whicheconomic activities bear the greatest burden ofenvironmental regulation. Canada’s EPEaccounts show that five industries account fornearly 80% of all EPE: Mining, Pulp and paper,

Primary metals, Petroleum refining, and Energyutilities. Canada also provides a breakdown byprovince. The industrial and geographicdisaggregation allows policy makers to identifythe industries and communities that would bemost affected by new environmental policies,and to design measures to assist them ifnecessary.

There has been some criticism that the EPEaccounts focus too much on the expenditureside which emphasizes the extra costs imposedby environmental regulation. Possibilities forrevenue and cost savings throughimplementation of process-integratedenvironmental measures have not received thesame degree of attention. A Swedish EPE surveyincluded questions about cost savings andfound a large share of companies were engagedin cost-reducing or revenue-enhancingmeasures that were not covered by the standardEPE survey instrument (Johansson 1998). Ascompanies are adopting process-integrated,pollution prevention instead of pollutionabatement approaches, the conventional EPEaccounts are less useful in analyzing the

Na: Not available.

1. Includes only government environmental protection expenditures.

2. Figures for Korea are for 1995.

Source: Table adapted from Alfieri 1998.

Table 14. Summary indicators of environmental protection expenditures in 1992

Philippines1

Republic ofKorea

2Costa Rica1 Colombia

Total EPE/GDP na 1.72 na 1.08

Government EPE/GDP 0.37 0.79 0.39 0.37

Capital EPE/Total EPE 61 49.7 50 72.7

Current EPE/Total EPE 39 50.3 50 27.3

EPE by environmental

media/Total EPE

Forest and non-

forest ecosystems

na na 28 2.7

Air Na 18.3 2.4

Water and soil Na 46.9 14 30

Waste Na 30.1 36 19.4


Environmental Protection and Resource Management Accounts

economic impact of environmental regulation,or the likely response to changes in regulation.

Resource management expenditures

This component of the accounts includesspending on resource management. The use ofthis account for improving resourcemanagement was discussed in section 3, theasset accounts. At present, these accounts havebeen constructed by a few European countries,Chile and Costa Rica.

Environmental protection industry

While EPE have imposed substantial costs, theyhave also created opportunities: entirely newindustries have arisen to fill the need forenvironmental services. The second part ofthese accounts provide a clear description ofthis industry: its contribution to GDP, toemployment, and to exports. For somecountries, the environmental services industryhas become an important exporter, while othercountries are large importers of these services.For example, in France, the environmentalservices industry accounted for 2.3% of GDPand 1.4% of employment on 1997. More thanhalf the employment was in solid waste andwaste water management (Desaulty and Templé1999).

Environmental and resource taxes

The third part of the accounts includes taxesand other fees collected by government forpollution emissions and for resource use, suchas levies on minerals, forestry or fisheries.Environmental taxes and subsidies areimportant policy instruments for achievingsustainability. Many European countries areexploring the possibility of substituting greentaxes for other forms of taxes to achieve a“double dividend” (revenues + a cleanerenvironment). The tax component of the EPEaccount can be very useful in assessing whetherthe tax regime provides incentives ordisincentives for sustainable development, and

whether taxes truly reflect the polluter paysprinciple that has been adopted by manycountries. Taxes on specific natural resources,and their use in resource management werediscussed in the section on asset accounts.

Eurostat has compiled a time series ofenvironmental taxes for its 15 member countries(Steurer and others 2000). As a share of total taxrevenue, environmental taxes constitute a smallbut increasingly significant tax, growing from6.7% of total tax revenue in 1980 to 7.6% in 1997.Among the different environmental taxes—taxes on energy, transport, pollution, andresources—energy taxes dominate and presentlyaccount for about three-quarters ofenvironmental taxes.

The use of environmental taxes has beenincreasing, yet there may be subsidies to someindustries established for other policy purposes(e.g., competitiveness, regional development,employment) that conflict with the intent of thegreen taxes and sustainable development. Todetermine whether environmental taxes havebeen successful in implementing the polluterpays principle, further analysis of the EPE taxaccounts is needed. The Swedish EnvironmentalProtection Agency undertook such a study.Environmental taxes rose from 49.7 billion SEKin 1993 to 61.6 billion SEK in 1998, and aredominated by energy taxes, which includecarbon taxes (Table 15).

The study defined subsidies to include bothdirect subsidies as well as tax subsidies.Potentially harmful direct subsidies include, forexample, support to home building, agriculture,forest road building, fisheries, and reindeerhusbandry. Tax subsidies include lower thannormal taxes on carbon and energy forindustries such as Transportation, electricityand gas works, and mining. Linked with thephysical accounts for energy, the tax accountsshow the extent to which the “polluter pays”principle is being followed.



A summary of all the taxes and environmentallyharmful subsidies by industry is shown inFigure 10. In some industries, theenvironmentally harmful subsidies received

outweigh the environmental taxes paid. Thelargest environmentally harmful subsidies go toreal estate, mainly interest rate subsidies forhousing construction. Agriculture, forestry and

Table 15. Environmental taxes in Sweden, 1993–98

(millions of SEK in current prices)

Taxes on: 1993 1994 1995 1996 1997 1998

Energy 39,017 42,043 44,161 49,733 49,352 52,652

Pollution 582 566 682 753 551 508

Transport 8,119 5,852 5,798 6,721 6,451 6,336

Resources - - - 70 131 142

Total 49,711 50,455 52,636 59,273 58,482 61,636

-40000 -30000 -20000 -10000 0 10000 20000 30000 40000

01-05 Agriculture, fishing & forestry

10-14 Mining & quarrying

15-37 Manufacturing

40-41 Elect., gas & water

45 Construction

50-52, 55 Wholesale, retail trade

60-64 Transp. & communication

65-67 Financial intermediation

70-99 Other

Private consumption

Public consumption

Ind

us

trie

s

SEK millon

Total environmental harmful subsidies

Total environmental taxes

Figure 10. Total environmental taxes and total direct subsidies, by industries and final

demand, 1995 (SEK million)

Source: Taken from Table 11 of Sjölin, M. and A. Wadeskog 2000.


Environmental Protection and Resource Management Accounts

fishing also receive more harmful subsidies thanthe environmental taxes they pay. Apparentconflicts between environmental taxes andenvironmentally harmful subsidies reflectdifferent policy objectives. The accounts makeclear what this relationship is and can provide abasis for discussion with industry about howbest to achieve environmental objectives

5.2 Policy analysis

There are a number of areas where theenvironmental protection and resourcemanagement accounts could be used for policyanalysis, such as:

• Economic impact of environmentalregulation

• Economic impact of environmental taxes• Assessment of the costs of regulation

relative to their benefits• Impact of recycling and reuse of materials.

No developing countries have yet used theirEPE accounts for policy analysis. Some analysishas been done in industrialized countries, likethe United States and the Netherlands, butgiven the problems of interpretation of EPEdata, additional information is often required.


6Economy-Wide Indicators ofSustainable Development

Many practitioners have searched for a way tomeasure sustainability either by revisingconventional macro indicators or by producingnew ones in physical units (listed in Table 2,section 2.6). Aggregate environmental themeindicators measured in physical units arederived from the NAMEA component of theSEEA. The physical indicators are meant to beused in conjunction with conventional economicindicators to assess environmental health andeconomic progress. A number of differentrevised environmental monetary aggregateshave been calculated by different countries; allare discussed in the revised SEEA. At this time,there is no consensus over which indicators touse. Rather, since each indicator serves asomewhat different policy purpose, the choiceof indicator depends on the policy question.

Two rather different questions can be posed bypolicy-makers: first, is the current level ofnational income sustainable? Second, what levelof income would be sustainable? This sectionbriefly reviews the experience with economy-wide physical indicators, then discusses theenvironmentally-adjusted macroeconomicindicators.

6.1 Physical indicators of macro-levelperformance

Section 2 discussed the use of macroeconomicindicators measured in physical units either asan alternative to monetary indicators (forassessing ecological, or strong, concepts ofsustainability), or to be used in conjunction withmonetary aggregates, providing environmental

information that the monetary aggregates maynot be able to capture fully. The NAMEAprovides physical macroeconomic indicators formajor environmental policy themes: climatechange, acidification of the atmosphere,eutrophication of water bodies, and solid waste.The trend of these indicators for theNetherlands are shown in Figure 11.

The indicators can be compared to a nationalstandard—such as the target level ofgreenhouse gas emissions—to assesssustainability. A national standard forgreenhouse gas emissions, set for example, interms of a country’s target under the KyotoProtocol can be useful. It may not be easy toassess some themes, such as eutrophicationwhich may have a more local impact, in termsof a national standard. The NAMEA does notprovide a single-valued indicator whichaggregates across all themes.

The Material Flow accounts (MFA) provideanother set of physical macroeconomicindicators, of which the most widely known istotal material requirements (TMR). TMR sumsall the material use in an economy by weight; itspurpose, like the monetary aggregates, is toprovide a single-valued indicator to measuredematerialization, the decoupling of economicgrowth from material use, a conceptpopularized by the Factor 4 movement. TheWorld Resources Institute study of MFA for fiveindustrialized countries shows significant de-coupling: since 1975, the material intensity ofGDP in all five countries has declined by 20-



40% (Figure 12). This has been the result ofefforts to reduce the volume of solid waste, andthe shift away from energy- and material-intensive industries toward knowledge-basedand service industries. Per capita materialintensity has not declined in most countriesover this time period; only Germany showed adecline (6%).

The limitations of these indicators ofdematerialization, discussed in section 2,restricts their use for assessing sustainability.Indicators created for certain categories ofmaterials whose environmental impacts aremore similar may be more useful, such as theNAMEA theme indicators. Nevertheless, the

relation between these indicators and totaloutput, or trends over time can provide insightinto whether countries are working towardsoverall goals of reducing their impact on theenvironment.

6.2 Environmentally-adjusted NDP andrelated indicators

The most well-known indicator in this categoryis the environmentally-adjusted NDP, eaNDP(sometimes GDP has been used instead). Thisindicator was calculated in early work onenvironmental accounting by Repetto and hiscolleagues as a way of focusing the attention ofpolicy-makers on the importance ofenvironmental degradation and depletion of

Source: Statistics Netherlands, unpublished data.

Figure 11. Index of macro-indicators for economic and environmental performance, Netherlands,

1987 to 1998 (1987 = 1.00)

70

80

90

100

110

120

130

140

150

1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Green House Gas

Acidification

Eutrophication

Waste (exc. re-use)

GDP


Economy-Wide Indicators of Sustainable Development

natural capital. Repetto’s work in Indonesia (onpetroleum, forests, land degradation) and CostaRica (on forests, fisheries, land degradation)was followed by similar pilot studies in Papua-New Guinea, and Mexico sponsored by the UNand the World Bank.

More recently, partially adjusted eaNDPs havebeen calculated for a number of countries,including Japan, Korea, the Philippines,Sweden, and Germany. The results aresummarized in Table 16. The great differencesamong countries in terms of the types ofcoverage and how the maintenance costapproach was implemented make it impossibleto directly compare results across countries.Korea, for example, assumed the sameabatement costs in all industries, whereas theother countries estimated industry-specificabatement costs.

Sweden’s eaNDP, calledGenuine Income, showsthe least change fromconventional NDP, only0.6%. One reason for thisvery low figure, despitesubtracting someenvironmental protectionexpenditures whichother countries did notdo, is that it measuresonly environmentaldegradation from sulfurand nitrogen. Swedenalso excludeddegradation not alreadyincluded in conventionalmeasures of NDP,whereas other studies,notably the Korean andPhilippine, did notexplicitly address theissue of potential double-counting. Theadjustment for Japan andGermany are ratherlarge, mainly because

they include the estimated cost of reducingcarbon emissions (and, for Japan, CFCs). Theother studies did not address these globalpollutants.

While Sweden and Germany have calculatedthese measures, they have not become part ofthe official statistics of these countries becauseof the problems with this measure discussed insection 2. The United States has expressed aninterest in measuring some form of eaNDP, butbased on the damage-cost approach andincorporating estimates of the benefits ofenvironmental services not currently includedin the accounts (Nordhaus and Kokkelenberg1999). However, the US has not been able toproceed due to lack of funding.

Figure 12. Percentage change in material use in five industrialized

countries, 1975 to 1996

Notes: Material intensity calculated as Domestic Processed Output/GDP.

Per capita material intensity calculated as Domestic Processed Output/Population.

Domestic Processed Output = Domestic extraction + Imports – Net additions to stock – Exports.

Source: Based on (WRI 2000, Table 2, page 20).

-50

-40

-30

-20

-10

0

10

20

Austria Germany Japan Netherlands United States

pe

rce

nta

ge

ch

an

ge

be

twe

en

19

75

an

d1

99

6

Material intensity of GDP

Per capita material intensity



Genuine savings

Genuine Savings, which adjusts measures ofwealth (actually, changes in wealth) rather thanincome has been calculated for many countries.A comparison of gross domestic savings togenuine savings for different regions of theworld in 1997 is given in Table 17, based onestimates for human capital represented byexpenditures on education, depletion of threemajor resources—energy, minerals, andforests—and environmental damagerepresented by carbon dioxide emissions. Crudeassumptions were made in order to calculatethis indicator for all countries. In all instances,

genuine savings is less than gross domesticsavings, but there is great variation amongregions. In the Middle-East and North Africa,genuine savings are actually negative. As theauthor notes, a negative figure for any one yeardoes not indicate that development isunsustainable—only if the negative trendpersists over time does this becomeunsustainable.

6.3 Modeling approaches to macro-economic indicators

Environmentally-adjusted NDP has beencriticized for combining actual transactions

Table 16. eaNDP as percentage of NDP in selected countries

Note: See text for a more detailed description of coverage and valuation methods.

Source: Japan—(Oda and others 1998), Korea—(Korea Environment Institute and others 1998), Philippines—(NSCB,

1998, 2000), Germany—(Bartelmus and Vesper 2000), Sweden—(Skanberg 2000).

Country and time period

Affect onmacroeconomic

aggregates Coverage Valuation method

Japan, 1990 NDP reduced 2.4% Depletion of minerals Net price method

Degradation of land, air,

water, including CO2

and CFCs Maintenance cost

Korea, 1985-1992

NDP reduced 4.1-

2.6% Depletion of minerals Net price method


water Maintenance cost

Philippines, 1988 to

1996

Depletion of forests, fish,

minerals Net price method


water Maintenance cost

Germany NDP reduced 3% Depletion of minerals Net price method


water including CO2 Maintenance cost

Sweden, 1993 and

1997 NDP reduced 0.6% Depletion of minerals Economic depreciation

Environmental damage

due to SOX, NOX only Damage cost

Environmental protection

expenditures Market cost



(conventional GDP and NDP) with hypotheticalvalues (monetary value of environmentaldegradation). The response to this criticism ledto the construction of a new set of indicatorsthat seek to estimate what sustainable nationalincome would be if the economy had to changeto meet the environmental constraints. Twomajor approaches were developed Hueting’sSNI and the geNDP.

Hueting’s SNI is described in section 2 as themaximum income that can be sustained withouttechnological development (excluding the use ofnon-renewable resources). Using a static,applied general equilibrium model, SNI hasbeen calculated for the Netherlands in 1990(Verbruggen and others 2000). The authorsfound that enormous changes would have tooccur in order to fulfill the sustainabilitystandards in the short term: SNI is 56% lowerthan national income in the base year (Table 18).Household consumption declines by 49%,government consumption by 69% and netinvestment by 79%. Both exports and importsfall by nearly two-thirds. Production in all

industries falls. Revenues from pollution taxesare so high that they replace all other taxes.Taxes exceed government consumption and areredistributed as lump sum payments tohouseholds to pay for consumption. Thepurpose of the SNI is not to provide policy-makers with a goal for national income as such,but to indicate the difference between currentincome and sustainable income. If this exercisewere updated, with new technologiesintroduced into the model as they becomeavailable, Hueting’s SNI would indicatewhether the economy is becoming more or lesssustainable over time.

An alternative approach, the geNDP, estimatesnational income looking into a hypotheticalfuture in which economic development mustmeet certain environmental standards. Theimpact on the economy is estimated byinternalizing the costs of reducingenvironmental degradation. The purpose of thisapproach is to provide policy-makers withguidance about the likely impacts of alternativedevelopment paths and the instruments for

Table 17. Genuine saving as percent of GDP, 1997

Source: Adapted from World Development Indicators (World Bank 1999), quoted in (Hamilton 2000, table 6.1)

Grossdomestic

savings

Consumptionof fixedcapital

Netdomestic

savingsEducation

expenditureEnergy

depletionMineral

depletion

Netforest

depletion

Carbondioxide

damage

Genuinedomestic

savings

World 22.2 11.7 10.5 5.0 1.2 0.1 0.1 0.4 13.6

Low income 17.0 8.0 9.1 3.4 4.2 0.6 1.8 1.2 4.8

Middle income 26.2 9.2 17.0 3.5 3.8 0.5 0.2 1.1 15.0

High income 21.4 12.4 9.0 5.3 0.5 0.0 0.0 0.3 13.5

East Asia &

Pacific

38.3 6.9 31.4 2.1 0.9 0.5 0.7 1.7 29.7

Europe &

Central Asia

21.4 13.7 7.9 4.2 4.9 0.1 0.0 1.6 5.6

Latin America

& Carib.

20.5 8.3 12.2 3.6 2.7 0.7 0.0 0.3 12.1

Middle East &

N. Africa

24.1 8.8 15.3 5.2 19.7 0.1 0.0 0.9 -0.3

South Asia 18.2 9.1 9.1 3.8 2.1 0.4 2.0 1.3 7.1

Sub-Saharan

Africa

16.8 9.1 7.8 4.5 5.9 1.4 0.5 0.9 3.4



achieving them. In these models, technologyand other model parameters are not alwaysrestricted to what is currently available.Estimates for the Netherlands were carried outby (De Boer and others 1994). A similar studywas carried out by the Swedish NationalInstitute of Economic Research (NIER, 2000)focusing specifically on CO2 emissions.

NIER carried out simulations in order to informthe Climate Committee of Sweden about thelikely macroeconomic impacts of achievingdifferent levels of reduction of CO2 emissions.Three scenarios were constructed, specifyingdifferent percentage reductions for CO2. Underthe Kyoto Protocol, Sweden agreed to stabilizeCO2 emissions at a level 4 percent higher thanits emissions in 1990. However, there is also awidespread view in Sweden that CO2 emissionsshould be reduced even further, so thealternative scenarios included two morerestrictive target levels: 2 and 8 percent lowerthan the emissions in 1990; the last one beingthe same as the commitment for EU as a whole.

For these simulations, a static generalequilibrium model was used with 18 productionsectors including public sector, one householdsector, and 6 types of energy inputs (from theenergy accounts). The model had fourmechanisms to reduce CO2 emissions in themodel: enhance energy efficiency, change to lessuse of coal fuels, lower production in coal-intensive industries (and expansion ofindustries which are less coal intensive) and,finally, to cut down on production (andconsumption).

The baseline scenario outlines a “business-as-usual,” probable growth path for the economyup to 2010 with no restrictions on CO2emissions. In this case, the emissions would riseto approximately 65 Mton, which is about 17percent higher than level in 1990 (56 Mton). Tokeep emissions within 4 percent of emissions in1990 thus requires a large increase of the CO2emission tax. The tax rate 1997 was (withexceptions for energy intensive export sectors2)

Base 1990 SNI Change (%)

National Income 457 201 -56

Private households consumption 314 159 -49

Government consumption 75 23 -69

Net investments 51 11 -79

Trade balance 16 8 -53

Exports 229 80 -65

Imports -213 -72 -66

National Product 457 201 -56

Agricultural production 15 4 -76

Industrial production 113 19 -83

Services production 242 37 -85

Taxes on production 88 0 -100

Pollution rights 0 165

Double counting 0 -24

Table 18. Hueting’s Sustainable National Income (Billions of guilders)

Note: SNI Variant 2b: constant trade shares; new equilibrium prices.

Source: Based on Table 7.5 of Verbruggen and others 2000.



370 SEK per ton CO2. In the +4 scenariorepresenting the Kyoto Protocol, the tax wouldhave to be increased to 820 SEK per ton CO2.The changes in the economy would reduce GDPslightly, by 0.3% under the +4 scenario, and by0.4% and 0.6% under the –2% and –8%scenarios, respectively (Table 19). Consumptionis not much affected except by the –8% scenario.All other macroeconomic variables areaffected—investment, trade and real income.

As a result of this study, the Climate Committeerecommended that the emissions of greenhousegases should, as a mean value for the period2003 o 2012, be 2 % lower than emissions in1990, counted as tons of carbon dioxideequivalent. The committee also suggested as along-term objective to reduce greenhouse gasesto 50 % of 1990 levels by 2050, but no politicaldecision has yet been taken.

The models used to estimate a sustainableeconomy, whether Hueting’s SNI or geNDP,vary not only in terms of the issues they addressand the assumptions they make, but also in

terms of the type of modelused. Many of the SNIexercises have used generalequilibrium models, butthese models are typicallystatic, that is, they do notdescribe the dynamic pathan economy would followto achieve the transitionfrom the present economyto a sustainable one. Meyerand Ewerhart (1998)addressed the greenhousegas issue for the German

economy with a dynamic econometric input-output model. The model is more disaggregatedby industry and uses the sale of carbon emissionpermits as a means to achieve the target levelsof emission reduction by 2005 (from 5% to 30%lower than Germany’s emissions in 1990). A 5%reduction of CO2 emissions resulted in virtuallyno change in GDP, relative to the baselinescenario, but a 30% reduction would reduceGDP by 3%, which is still rather small. Thedynamic adjustment of the economy over time(1996 to 2005) is not smooth, suggesting thatsuch a model provides important information topolicy-makers.

Because of the many assumptions thatpractitioners must make in constructing theseindicators, the wide range of environmentalfactors which practitioners may choose toinclude or ignore, and the differentmethodologies which have been used, nostudies have produced indicators that arecomparable across countries.

Table 19. Macroeconomic effects of measures to reduce CO2 in Sweden

Source: S. Ahlroth 2000. “geGDP simulations in Sweden: Simulations of tax policy for CO2

reductions.” Unpublished paper of Swedish National Institute of Economic Research.

Percent change compared to

baseline scenario, 2010

+4% Scenario -2% Scenario -8% Scenario

GNP -0,3 -0,4 -0,6

Private consumption 0,0 -0,1 -0,2

Public consumption 0,0 0,0 0,0

Investment -0,4 -0,7 -1,1

Exports -0,6 -1,0 -1,3

Imports -0,5 -0,7 -0,9

Real income -0,2 -0,4 -0,5


7 Concluding Comments

Much of the use of environmental accounts hasbeen in industrialized countries, especiallyEurope, Australia, and Canada. The assetaccounts are compiled by most countries, butnot used very much in assessing sustainability.The flow accounts are widely used, both for theconstruction of indicators and as inputs topolicy modeling. The construction of monetaryenvironmental macro indicators is quite limited,and it is not clear that these indicators havebeen much used.

There are, in addition, four main observationsregarding how useful environmental accountsare for policy:1. Although some countries are using the

environmental accounts quite actively, theaccounts are still underutilized, especiallyin developing countries

2. No country has truly comprehensiveenvironmental accounts

3. International comparisons are important,but not yet possible because of differencesin methodology, coverage, environmentalstandards, and other factors

4. For a country to fully assess itsenvironmental impact, it must have• Accounts for the transboundary

movement into and out of the countryof pollutants via air and water

• Accounts for its major trading partnersto calculate the pollution and materialcontent of products that it imports.

Underutilization of accounts

The asset accounts have been used to monitorsustainability in various ways, but many

countries have not exploited their full potentialto monitor characteristics of wealth and changesin wealth over time. This may be due to thelack of emphasis on conventional asset accountsand measures of wealth. The lack of aconsensus in the revised SEEA about a methodfor measuring the cost of depletion is also adeterrent. The asset accounts could also bemore widely used to assist in resourcemanagement. Even simple analysis, such ascomparison of rent to the taxes on rent and thecosts of resource management is not routinelycarried out in countries that compile assetaccounts for natural capital.

The flow accounts are more widely used, bothfor the construction of indicators, environmentalprofiles, and analysis. There is considerableoverlap between the SEEA and thesustainability indicators proposed by the UnitedNations, OECD, and other organizations.Tighter links among these different approachescould be useful.

Comprehensive environmental accounts

Sustainability can only be measured if all assetsare included. Including natural capital as partof a country’s wealth is an important steptoward better measure of sustainability.However, most countries do not include allenvironmental capital, or all the environmentalgoods and services that natural capital provides.For example, asset values most often reflect onlythe value of the major commercial product, e.g.,timber value of forests.



International comparability

International comparisons are extremely usefulfor countries in assessing their resourcemanagement. The comparison of wateraccounts in southern Africa or theenvironmental damage costs in Europe, forexample, are extremely helpful for policy. Sofar, the comparison of accounts and of theresulting indicators across countries is notgenerally possible because of the wide range ofdefinitions, coverage and methodologies usedby different countries. Monetary accounts maydiverge even more than physical accountsbecause of the range of different valuationmethodologies that could be used,environmental standards, and otherassumptions necessary for valuation. With theexception of the Genuine Savings indicator, ithas not been possible to compare monetaryenvironmental macro indicators acrosscountries.

Full assessment of environmental impact

Several studies in Europe have shown that thequantities of pollution exported and importedvia air and water are very large; withoutaccurate information about these quantities, theuse of environmental accounts for policy will belimited. Similarly, substantial pollution andresources are embodied in international trade.The Swedish study showed that environmentalcoefficients can diverge substantially amongcountries, and that a proper assessment of theenvironmental impact of a country’s importscan only be made with information about theenvironmental coefficients of one’s tradingpartner, from the partner’s environmentalaccounts.

In addition, management of global or regionalenvironmental problems, whether climatechange or acidification, require comparableenvironmental accounts for each country.

Progress toward internationalcomparability

Regional organizations and projects provide agood forum for developing comparableenvironmental accounts. The European Unionhas led several efforts, compiling oil and naturalgas accounts for five member countries, andforest accounts for three countries. It hassupported the GREENSTAMP (GREENedNational STAtistical and Modeling Procedures)Project to work on developing environmentalaccounts and studying their potentialapplications in the EU. Eurostat has alsosupported the Green Accounting ResearchProject (GARP) which attempts to compilecomparable monetary accounts based on thedamage cost approach, includingtransboundary transport, across 15 EUcountries.

A regional program to develop environmentalaccounts began in the late 1990’s in southernAfrica, focusing on Namibia, Botswana, andSouth Africa with support from internationaldonors. As a regional undertaking, the projecthas emphasized coming to an agreement aboutcommon methods in order to constructcomparable accounts among the countries. Theformation of the Manila Forum, a group ofsoutheast Asian countries promotingdevelopment of environmental accounts, alsoprovides a good opportunity for buildingcomparable environmental accounts in thisregion.


Notes

1. Data for the first two methods wereobtained for 1995; data for the third methodwere only available for 1993, so the resultsare not directly comparable with resultsfrom the other two methods.

2. The manufacturing industry (= pulp andpaper, chemical, refineries, iron and steel,engineering, and other manufacturingindustries in the Excel figures) pays 50% ofthe CO2 tax, i.e., 0.185 SEK/kg CO2.


References

Ahlroth, S. 2000a. Correcting NDP for SO2 andNOx emissions: Implementation of a theoreticalmodel in practice. National Institute for EconomicResearch (Konjunkturinstitutet).

_____. 2000b. “geGDP simulations in Sweden:Simulations of tax policy for CO2 reductions.”Paper of Swedish National Institute ofEconomic Research. Processed.

_____. 2000c. Data about costs of environmentaldamage in Sweden. NIER: Stockholm.Processed.

Ahmad, Y.J., S. El Serafy, and E. Lutz, eds. 1989.Environmental Accounting for SustainableDevelopment. The World Bank, Washington.

Alfieri, A. 1998. “Environmental ProtectionExpenditures: The Experience of DevelopingCountries.” Paper presented at 5th AnnualMeeting of the London Group on Environ-mental Accounting. Fontevraud, France, 25-29May.

Asheim, G., and W. Buchholz (forthcoming).“The Hartwick Rule: Myths and Facts.”Unpublished paper, Department of Economics,University of Oslo.

Australian Bureau of Statistics. 2000. WaterAccounts for Australia, 1993–93 to 1996–97. ABS:Canberra.

_____. 1999. Consolidated Balance Sheet. ABS:Canberra.

Bartelmus, P., E. Lutz, and S. Schweinfest. 1992.Integrated Environmental and EconomicAccounting: A Case Study for Papua-New Guinea.World Bank Environmental Working Paper No.54.

Bartelmus P., and A. Vesper. 2000. “Greenaccounting and material flow analysis:Alternatives or complements?” WuppertalInstitute Paper No. 106.

Batcagan, S. 2000. “Pricing of grassland resourcein the Philippines: Rent, grassland degradationand rehabilitation, and alternative land uses.”Paper presented at the International Workshopon Environmental and Economic Accounting,18-22 September, Manila.

Botswana, Ministry of Finance andDevelopment Planning. 1997. Mid-Term Reviewof NDP7. Gaborone.

Cleveland, C., and M. Ruth. 1999. “Indicators ofdematerialization and materials intensity ofuse.” Journal of Industrial Ecology 2 (3): 15-50.

Davis, G., and D. Moore. 2000. “Valuing mineralstocks and depletion in green national incomeaccounts.” Environmental and DevelopmentEconomics Vol. 5, parts 1 and 2, February andMay, pp. 109-128.

De Boer, B., M. de Haan, and M. Voogt. 1994.“What would Net Domestic Product have beenin an environmentally sustainable economy?”

The word processed describes informally reproduced works that may not be commonly available through libraries.



Presented in Papers and Proceedings of theMeeting on National Accounts and theEnvironment, 16–18 March, London.

de Haan, M., S.J. Keuning, and P. Bosch. 1993.Integrating Indicators in a National AccountingMatrix Including Environmental Accounts.(NAMEA), NA 060, 1993, Voorburg: CentralBureau of Statistics.

Desaulty, D., and P. Templé. 1999. “In 1997,France spent 145 billion francs onenvironmental protection.” Presented at Jointmeeting of the Working Party EconomicAccounts for the Environment, 6-7 December.Eurostat, Luxembourg.

Das Gupta, P., and K-G. Mäler. 2000. “Netnational product, wealth, and social well-being.” Environmental and DevelopmentEconomics Vol. 5, parts 1 and 2, February andMay, pp.69-94.

Ecotec. 1999. Who Gains from the Climate ChangeLevy? Final report to WWF-UK. Ecotec Researchand Consulting, Ltd. London, Birmingham,Bruxelles, and Madrid.

El Serafy, S. 1989. “The Proper Calculation ofIncome from Depletable Natural Resources.” InAhmad.

ENRAP. 1999. ENRAP-SHELF CD-ROM. CDcontaining all accounts and technical reportsfrom the Philippine Environmental and NaturalResource Accounting Project. Manila.

Eurostat. 2000a. Accounts for Sub-Soil Assets-Results of pilot studies in European countries.Eurostat, Luxembourg.

_____. 2000b. Valuation of European Forests—Results of IEEAF Test Applications. Eurostat,Luxembourg.

Førsund, F. 1985. “Input-Output Models ,National Economic Models, and theEnvironment.” In A.V. Kneese and J. L.Sweeney, eds., Handbook for Natural Resource andEnergy Economics, Vol. 1, 325-341, ElsevierPublishing Co.: New York.

Gloria, Thomas. 2001. “A Dynamic Life-cycleApproach by Evaluating Structural EconomicSequences.” In Chapter 14 of A.M. Nielsen andB.P. Weidema, eds., Input/Output Analysis—Shortcuts to Life Cycle Data? Copenhagen:Danish Ministry of Environment and Energy.

Hamilton, K. 2000. “Greening the NationalAccounts: Formal Models and PracticalMeasurement.” In J. Proops and S. Simon, eds.,Greening the Accounts. Cheltenham, UK: EdwardElgar Publishers.

Hamilton, K., and M. Clemens. 1999. “GenuineSavings Rates in Developing Countries.” WorldBank Economic Review 13:2, 333-56.

Hartwick, J.M. 1990. “Natural Resources,National Accounting and EconomicDepreciation.” Journal of Public Economics 43:291-304.

Hass, J., and K. Sorenson. 1998. Environmentalprofiles and benchmarking of Norwegianindustries.” Economic Survey. January, pp. 28–37. Oslo: Statistics Norway.

Hassan, R. 2000. “Improved measure of thecontribution of cultivated forests to nationalincome and wealth in South Africa.”Environmental and Development Economics Vol. 5,parts 1 and 2, February and May, pp. 157-176.

Hassan, R. 1998. “Evaluation of the economy-wide impacts of the new water policy changesin South Africa: A SAM approach.” Paperpresented at the World Congress of Environ-mental and Resource Economists, 25-27 June.Venice, Italy.


References

Hellsten, E., S. Ribacke, and G. Wickbom. 1999.“SWEEA—Swedish environmental andeconomic accounts.” Structural Change andEconomic Dynamics 10 (1) 39-72.

Hicks, J.R. 1946. Value and Capital. 2nd Edition,Oxford University Press, Oxford.

Hueting, R. 1991. “Correcting national incomefor environmental losses: Toward a practicalsolution for theoretical dilemma.” In R.Costanza, ed., Ecological Economics, The Scienceand Management of Sustainability. New York:Columbia University Press.

Johansson, U. 2000. “Environmental protectionexpenditure in industry in 1997—Results of theSwedish pilot study.” Stockholm: StatisticsSweden, Environment Statistics.

Komen, M., and J. Peerlings. 1999. “Energytaxes in the Netherlands: What are thedividends?” Environment and Resource EconomicsVol. 14 pp. 243-268.

Korea Environment Institute, UNDP, andUNSD. 1998. Pilot compilation ofenvironmental-economic accounts. KoreaEnvironmental Institute: Seoul.

Keuning, S., J. van Dalen, and M. de Haan. 1999.“The Netherlands’ NAMEA: Presentation,usage and future extensions.” Structural Changeand Economic Dynamics 10 (1) 15-37.

Kunte, A., K. Hamilton, J. Dixon, and M.Clemens. 1998. “Estimating National Wealth:Methodology and Results.” EnvironmentDepartment Papers, Environmental EconomicsSeries No. 57, Washington: The World Bank.

Lange, G. 1997. “Strategic planning forsustainable development in Indonesia usingnatural resource accounts.” In J. van den Berghand J. van der Straaten, eds., Economy and

Ecosystems in Change: Analytical and HistoricalApproaches. Aldershott, UK: Edward ElgarPublishing.

_____. 2000a. “The contribution of minerals tosustainable economic development inBotswana.” Report to the Botswana NaturalResource Accounting Programme, NationalConservation Strategy Agency and Ministry ofFinance, Central Statistics Office. Gaborone,Botswana.

_____. 2000b. “The Use and Policy Applicationsof the Philippine System of Environmental andNatural Resource Accounts.” Report for thePhilippines National Statistical CoordinatingBoard. Manila. March 2000.

_____. (forthcoming). “Fisheries accounting inNamibia.” In C. Perrings and J. Vincent, eds.,Natural Resource Accounting and EconomicDevelopment: Theory and Practice. Cambridge:Cambridge University Press.

Lange, G., J. Arntzen, S. Kabaija, and M.Monamati. 2000. “Botswana’s Natural ResourceAccounts: The Case of Water.” Report to theBotswana Natural Resource AccountingProgramme, National Conservation StrategyAgency and Ministry of Finance, CentralStatistics Office. Gaborone, Botswana.

Lange, G., and R. Hassan. 1999. Natural resourceaccounting as a tool for sustainable macroeconomicpolicy: Applications in Southern Africa. PolicyBrief of IUCN Regional Office for SouthernAfrica (Harare, Zimbabwe, IUCN-ROSA).

Lange, G., J. MacGregor, S. Masirembu, and H.Suich. 2001. “Valuing water in Namibia: Twocase studies of agriculture.” Draft ResearchDiscussion Paper. Directorate of EnvironmentalAffairs, Ministry of Environment and Tourism:Windhoek, Namibia.



Leontief, W. 1970. “EnvironmentalRepercussions and the Economic Structure: anInput/Output Approach.” Review of Economicsand Statistics II, August 1970.

Mäler, K.-G. 1991. “National Accounts andEnvironmental Resources.” Environmental andResource Economics 1 : 1-15.

Manasan, R., M. de los Angeles, L. Rufo, A.Inocencio, D. Ramirez, L. Castro, and C. Rufo.1998. “Alternative policies to control PM10emissions from motor vehicles in MetroManila.” June 1998. ENRAP/DENR Uneditedversion for discussion. Table 15.

Manning, P. 2000. “Review of the distributiveaspects of Namibia’s fisheries policy.” NEPRUResearch Report. June. Namibian EconomicPolicy Research Unit: Windhoek, Namibia.

Matthews E., C. Amann, S. Bringezu, M.Fischer-Kowalski, W. Huettler, R. Kleijn, Y.Moriguchi, C. Ottke, E. Rodenburg, D. Rogich,H. Shandl, H. Schuetz, E. van der Voet, and H.Weisz. 2000. The weight of nations—Materialoutflows from industrial Economies. WorldResources Institute, Washington D.C., 125pp.

Markandaya, A., A. Hunt, and P. Mason. 2000.“Valuing damages for green accountingpurposes: The GARP II approach.” FEEMWorking Paper No. 93.2000. Available online athttp://www.feem.it.

Meyer, B., and G. Ewerhart. 1998. “Multisectoralpolicy modelling for environmental analysis.”In K. Uno and P. Bartelmus, eds., EnvironmentalAccounting in Theory and Practice. Dordrecht:Kluwer Publishers.

Meza, F., R. Jiliberto, F. Maldini, A. Magri, M.Alvarez-Arenas, M. García, S. Valenzuela, andL. Losarcos. 1999. Cuentas ambientales del RecursoAgua en Chile. (translation: Environmental

Accounts for Water in Chile). Documento deTrabajo Nº 11, Serie Economía Ambiental,Comisión Nacional del Medio Ambiente.

Miller, R., and P. Blair. 1985. Input-OutputAnalysis: Foundations and Extensions. Prentice-Hall, Inc: Englewood Cliffs, New Jersey.

Munksgaard, J. 2001. “Applications of IO-Models for LCA: Some experiences from theenergy area.” In A.M. Nielsen and B.P.Weidema, eds., Input/Output Analysis—Shortcutsto Life Cycle Data? Copenhagen: Danish Ministryof Environment and Energy.

NIER (National Institute for EconomicResearch). 2000a. Environmental impacts ofSwedish trade—Results of a pilot study. Stockholm.

NSCB (National Statistical Coordination Board).1998. The Philippine Asset Accounts: Forest,Fishery, Minerals, Land/Soil, and Water. Manila.

_____. 2000. Environmental Degradation Due toSelected Economic Activities. Manila.

Nielsen, A. M., and B.P. Weidema, eds. 2001.Input/Output Analysis—Shortcuts to Life CycleData? Copenhagen: Danish Ministry ofEnvironment and Energy.

Nordhaus, W., and E. Kokkelenberg. 1999.Nature’s Numbers. Washington, D.C.: NationalAcademy Press.

O’Connor, M. 2000. “Toward a typology ofenvironmetally-adjusted national sustainabilityindicators: Key concepts and policyapplications.” Paper prepard for the LondonGroup.

Oda, K, K. Arahara, N. Hirai, and H. Kubo.1998. “Japan: The system of integratedenvironmental and economic accounting(SEEA)—trial estimates and remaining issues.”


References

In K. Uno and P. Bartelmus, eds., EnvironmentalAccounting in Theory and Practice. Dordrecht:Kluwer Publishers.

Pearce, D.W., and G. Atkinson. 1993. “CapitalTheory and the Measurement of SustainableDevelopment: An Indicator of WeakSustainability.” Ecological Economics 8 103-8.

Pearce, D.W., K. Hamilton, and G. Atkinson.1996. “Measuring Sustainable Development:Progress on Indicators.” Environment andDevelopment Economics 1 85-101.

Pearson, P. 1989. “Proactive Energy-Environment Policy Strategies: A Role for Input-Output Analysis?” Environment and Planning A,21, 1329-1348.

Pezzey, J. 1989. “Economic Analysis ofSustainable Growth and SustainableDevelopment.” Environment Dept. Working PaperNo. 15. The World Bank. Washington, D.C.

Peskin, H., and E. Lutz. 1990. “A survey ofresource and environmental accounting inindustrialized countries.” Environment WorkingPaper No. 37. World Bank: Washington, D.C.

Rademacher, W. 1998. “Land use accounting—Pressure indicators for economic activity.” In K.Uno, and P. Bartelmus, eds., EnvironmentalAccounting in Theory and Practice. Dordrecht:Kluwer Publishers.

Repetto, R., W. Magrath, M. Wells, C. Beer, andF. Rossini. 1989. Wasting Assets: NaturalResources in the National Accounts. WorldResources Institute, Washington, D.C.

Ryan, L. 2000. “Adjusting the Australiannational accounts for depletion of subsoilassets—Some illustrative compilations.”National Accounts Research Paper. ABS:Canberra.

Saleh, K. 2000. “Compilation of environmentstatistics and economic accounts—TheIndonesian experience.” Paper presented at theInternational Workshop on Environmental andEconomic Accounting, 18-22 September, Manila.

Serroa da Motta, R., and C. Derraz do Amaral.2000. “Estimating timber depletion in theBrazilian Amazon.” Environment andDevelopment Economics Vol. 5, parts 1 and 2,February and May, pp. 129-142.

Sjölin, M., and A. Wadeskog. 2000.“Environmental taxes and environmentallyharmful subsidies.” Report prepared for DGEnvironment and EUROSTAT. StatisticsSweden: Stockholm.

Skånberg, K. 2000. A partially environmentallyadjusted net domestic product for Sweden 1993 and1997. National Institute for Economic Research.Stockholm.

Spangenberg, J.H., and others. 1999. MaterialFlow-Based Indicators in Environmental Reporting.EEA: Copenhagen.

Statistics Canada. 1997. Eco-connections: Linkingthe Environment and the Economy. StatisticsCanada: Ottawa.

_____. 2000. Agricultural land use and supply.Statistics Canada: Ottawa.

Statistics Norway. 2000. Data on resource rent,taxes and subsidies for petroleum, forestry, andfisheries in Norway. Statistics Norway: Oslo.Processed.

Steurer, A., T. Jaegers, and S. Todsen. 2000.“Environmental taxes in the EU.” Statistics inFocus. Eurostat, Luxemburg.

Stott, A., and R. Haines-Young. 1998. “Linkingland cover, intensity of use and botanicaldiversity in an accounting framework in theUK.” In K. Uno and P. Bartelmus, eds.,



Environmental Accounting in Theory and Practice.Dordrecht: Kluwer Publishers.

Structural Change and Economic Dynamics. 1999.Environmental extensions of national accounts:the NAMEA framework. 10 (1).

Tropical Science Center and World ResourcesInstitute. 1991. Accounts Overdue: NaturalResource Depreciation in Costa Rica. Washington,D.C.: World Resources Institute. 110pp.

Tjahjadi, B., D. Schäfer, W. Rademacher, and H.Höh. 1999. “Material and energy flowaccounting in Germany—Data base for applyingthe national accounting matrix includingenvironmental accounts concept.” StructuralChange and Economic Dynamics 10 (1): 73-97.

United Nations. 1993a. System of NationalAccounts. United Nations: New York.

_____. 1993b. Integrated Environmental andEconomic Accounting. Series F No. 61. UnitedNations: New York.

van Tongeren, J., S. Schweinfest, and E. Lutz.1991. Integrated Environmental and EconomicAccounting: A Case Study of Mexico. World BankEnvironment Working Paper No. 50.Washington, D.C.: The World Bank.

Vaze, P. 1999. “A NAMEA for the UK.”Structural Change and Economic Dynamics 10 (1):99-122.Verbruggen, H., R. Dellink, R. Gerlagh, and M.Hofkes. 2000. “Calculations of a sustainablenational income: Four variants.” In H.Verbruggen, ed., Final report on calculations of asustainable national income according to Hueting’smethodology. Institute for EnvironmentalStudies, Vrije Universiteit: Amsterdam, TheNetherlands.

Vincent, J. 1997. “Resource depletion andeconomic sustainability in Malaysia.”Environment and Development Economics Vol. 2,pp. 19-37.

_____. 1999. “Net accumulation of timberresources.” Review of Income and Wealth Series45, Number 2, June, pp. 251-262.

Vincent, J., J. Hartwick, and T. Panayotou. 1997.“Resource depletion and sustainability in smallopen economies.” Journal of EnvironmentalEconomics and Management Vol. 33, pp. 274-86.

Von Weisäcker, E.U., A. Lovins, and H. Lovins.1997. Factor Four: Doubling Wealth—HalvingResource Use. Earthscan: London.

WCED (World Commission on Environmentand Development). 1987. Our Common Future.New York: Oxford University Press.

Weitzman, M.L. 1976. “On the WelfareSignificance of National Product in a DynamicEconomy.” Quarterly Journal of Economics (90)156-62.

World Bank. 1997. Expanding the Measure ofWealth: Indicators of Environmentally SustainableDevelopment. Environmentally SustainableDevelopment Studies and Monographs SeriesNo. 17, Washington: The World Bank.

_____. 1999. World Development Indicators 1999.Washington: The World Bank.

Wright, M. 1997. “The use of mineral revenuesin Botswana: Super caution vs. pragmatism.”The Research Bulletin of the Bank of Botswana Vol.15, no. 1, pp 35-47.

Documents

Policy Applications of Environmental Accounting