Embed Size (px)
Citation preview
Report No. 14961
Lending for Electric Power inSub-Saharan AfricaOctober 16, 1995
Operations Evaluation Department
DCUni*t of the a.,
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Pub
lic D
iscl
osur
e A
utho
rized
Abbreviations and Acronyms
AFDB African Development BankARPP Annual Review of Portfolio PerformanceCDC Commonwealth Development CorporationCFL Caisse Fran,aise de DevelopmentCIDA Canadian International Development AgencyCIE Compagnie Ivoirienne d'ElectriciteCIPREL Compagnie Ivoirienne de Production d'ElectriciteDCGTx Direction et Control des Grandes TravauxDEH Swiss Development CooperationECG Electricity Corporation of GhanaECPC Electricity Consumption Per CapitaEDF Electricite de FranceEECI Energie Electrique de C6te d'lvoireEIB European Investment BankENELGUI Enterprise National d'Electricite de GuineeERR Economic Rate of ReturnESAL Energy Sector Adjustment ProgrammeESMAP Energy Sector Management Assistance ProgrammeESW Economic and Sector WorkGDP Gross Domestic ProductGTZ Deutsch Gesellshaft fur Technische ZuzammenarbeitIMF International Monetarv FundIPP Independent Power ProducerKFAED Kuwait Fund for Arab Economic DevelopmentKfW Kreditanstalt fur WiederautbauLAC Latin America and the CaribbeanLDC Less Developed CountriesLRMC Long Run Marginal CostNEPA National Electric Power AuthorityNORAD Norweigian Agency for Development CooperationODA Overseas Development AdministrationOECD Organization for Economic Cooperation and DevelopmentOED Operations Evaluation DepartmentOLADE Organizacion Latino-Americana de EnergiaOMS Operations Manual StatementPAR Performance Audit ReportPCR Project Completion ReportPMTF Portfolio Management Task ForceREGIDESO Regie de Distribution d'Eau et d'ElectriciteROR Rate of ReturnSAL Structural Adjustment LoanSAR Staff Appraisal ReportSAUR Societe Africaine UrbaineSECAL Sector Adjustment LoanSIDA Swedish International Development AgencySODECI Societe de Distribution d'Eau de C6te d'lvoireSOE State Owned EnterpriseSSA Sub-Saharan AfricaTA Technical AssistanceTANESCO Tanzania Electric Services CompanyTMWS Task Manager Work StationUNDP United Nations Development ProgramVALCO Volta Aluminum CompanyVRA Volta River AuthorityWDR World Development Report
THE WORLD BANKWashington, D.C. 20433
U.S.A.
Office of Director-GeneralOperations Evaluation
October 16, 1995
MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT
SUBJECT: Lending for Electric Power in Sub-Saharan Africa
This study invesbgates whether Sub-Saharan Afica (SSA) posed-and will continue to pose-a specialchallenge m the Banles lending for power sector development. It evaluates the performance of 41 power projectscompleted fromn 1978 to 1993. Forthese projects-US$3.3 billion in total-the Bank lent US$1.2 billion andmobilized US$1.4 billion in co-financing. While the SSA record in the execution of physical components was onaverage as good as in other regions, this was not the case for institutional development and policy reforms. Lowoperational efficiency and failure to recover costs have plagued the sector despite repeated injection of Bankfunds.
Extmal factors and public sector weakness are part of the explanation. Borrower ownership of policyand institotional improvements agreed with the Bank was often missing. Lapses in the coordination amongiternational donors also hurt. But, of even larger significance, process factors, under the control of Bankmnagement and staff, also played a role.
Sub-Saharan countries need to increase the commercial nature of public utilities and involve privateoperators in the sector. Thus, the new Bank policy for the sector is well adapted to SSA circumstances. Thereaization that reform is needed is growing in SSA. In a few cases, progress has been made in te form ofmanagment contracts based on a blend of local and offshore know-how. The study recommends that the Bankshould nurture borrower ownership for such changes-particularly for a regulatory framework that ensuresautonomy and adequate power prices for operators, public or private. It also recommends stategic alliances anda more efficient division of labor among international donors. But it also cautions that, more than in the past, thecountry assistance strategies should examine the justification of power lending and establish its coherence withoter assistance priorities. In particular, the Bank should avoid making large power loans unless the sector isoperatng reasonably well or is substantially on its way to reforn.
Attachment
1
Contents
Executive Summary ........................................... 3
1. Introduction .11
Study Objective .......................................... 1 IStudy Scope and Methodology .......................................... 1 IReport Outline .......................................... 12The Public Utilities Paradigm of the 70s and 80s .......................................... 12The Response to the Energy Crises .......................................... 12The 1992 Policy Revision .......................................... 13The Core Challenge for SSA Countries .......................................... 13
2. Review of Completed Projects .......................................... 17
Objectives and Components .......................................... 17Perfornance .......................................... 20Institutional Development and Technical Assistance .......................................... 21Compliance with Covenants ........................................... 23Environment and Resettlement .......................................... 23
3. Program Impact .......................................... 25
Least Cost Development .......................................... 25Efficiency and Institution Strengthening .......................................... 26Access by the Poor .......................................... 31Resource Mobilization .......................................... 32
4. Performance Factors .......................................... 35
External and Country Factors .......................................... 35Pitching Project Objectives .......................................... 36Obtaining Borrowers' Ownership .......................................... 37Optimizing Bank Involvement .......................................... 40Effectiveness of Technical Assistance .......................................... 42Co-Financing and Donor Coordination .......................................... 45
5. Implementing the New Policy .......................................... 47
The New Policy ............................ : 47Overview of Recent Projects ........................... 47Transparent Regulatory Process ........................... 48Commercialization and Corporatization ........................... 51Importing Services ........................... 51Commitment Lending ........................... 53Private Investment ........................... 55
This report was prepared by Alvaro Covarrubias (Task Manager). Guy Pr6noveau, Ren6 Ribiand Gary Wu (Consultants) made valuable contributions to it. Charles Strout and Loma Sibbliesprovided administrative support.
2
6. Conclusions and Recommendations .............................................................. 57
The Road Ahead: Immediate Agenda ........................................................ 57The New Frontier: Restructuring and Assets Privatization .................................................. 58Role of the Bank ........................................................ 61Recommendations ........................................................ 63
Annexes
Annex A: Projects in the Study Cohort .............................................................. 65Annex B: On-going Projects .............................................................. 67
Tables in Text
Table 2.1: Lending for Completed Infrastructure Projects in SS Africa Since 1978 ................... 17Table 2.2: Power Projects Completed (1978 - 93): Ratings by OED .......................................... 20Table 2.3: Compliance with Financial Covenants .............................................................. 23Table 3.1: Operating Indicators in the Early 1990's .............................................................. 27Table 3.2: Financial Indicators in the Early 1990's .............................................................. 29Table 3.3: Financing of Power Projects in Sub-Saharan Countries in US$ million (%) ............. 32
Figures
Figure 1.1: GDP per Capita vs. Electricity Consumption Per Capita .......................................... 15Figure 2.1: Power Projects .............................................................. 18Figure 2.2: Project Objectives .............................................................. 18Figure 2.3: Distribution of Physical Components in the Cohort .................................................. 18Figure 2.4: Distribution of Non-Physical Components in the Cohort .......................................... 19Figure 2.5: Project Economic Rate of Return (ERR): Frequency Distribution ........................... 21Figure 3.1: Average Revenue .............................................................. 28Figure 3.2: Completed Projects - Contributions by Major Co-Financing Institution .................. 33
Boxes
Box 3.1: Nigeria: Recovering Cash Costs and Depreciation ......................................... ............. 30Box 3.2: Tanzania: Changing Revenue Covenants .............................................................. 31Box 4.1: Ghana: A Case of Mixed Ownership of Power Projects ............................................... 39Box 4.2: Bank Involvement in Project Preparation .............................................................. 42Box 4.3: Good Practice on Twinning with Expatriates .............................................................. 43Box 5.1: Burundi: Separation of Responsibilities on Paper ........................................................ 49Box 5.2: Good and Bad Practices on Financial Reporting ........................................................... 50Box 5.3: Good Practice on Management Assistance and Staff Streamlining .............................. 52Box 5.4: Good Practice on Management Contract ("Affermage") ............................................... 53Box 5.5: Good Practice on Commercialization Objectives .......................................................... 54
3
Executive Summary
I . Why are Bank power projects less successful in Sub-Saharan Africa than in otherregions? To find answers, this study establishes the outcomes of the completed projects andassesses the Bank's and Borrowers' performance. It identifies country, project, and processfactors that have enhanced or inhibited progress. It also assesses the prospects for implementingthe principles of the Bank's new power sector policy and suggests approaches to improve theimpact of Bank intervention.
2. For the 26 SSA countries receiving Bank credits or loans for electric power, the averageGDP per capita was US$304 in 1994. Their economies grew at 5.5 percent a year during 1970-80,and stagnated until 1985 before plummeting to their recent lows. Their combined energy resourceendowment is substantial with numerous hydro sites, coal and uranium deposits, and large crudeoil and gas reserves. But they use little of this potential, consuming 31 TWh of electricity in1971, and 81 TWh in 1991. On a per capita basis, electricity consumption has risen and fallenwith GDP: from 134 kWh in 1971 to 200 kWh in 1991 after peaking at 247 kWh in 1980.Compare those figures with 580 kWh for China and 350 kWh for India.
3. Core challenge. Power supply should not, as a rule, lag behind or race ahead ofeconomic growth. Many LDCs have tried to race ahead by setting unduly low power prices-hoping to spur their economic growth. But this has proven unaffordable and ineffective becauseelectrification programs do not significantly affect economic growth in the absence of other
development prerequisites.I In SSA as elsewhere, the challenge is for the power sector to meeteffective demand at least cost without (i) adding to public deficits, and/or (ii) competing forscarce public funds for such social priorities as education and health.
4. Public utilities paradigm of the 70s and 80s. Most power utilities benefiting from Bankloans have been state-owned enterprises, and all projects completed by 1993 were designed underthe Bank Operational Manual Statement 3.72 for public utilities projects, issued in 1978. OMS3.72 sets out broad policy objectives in accord with the public utilities paradigm prevailing at thattime. The Bank sought to help (i) provide power service on the basis of least cost developmentprograms, (ii) strengthen the sector's institutions and improve their efficiency, (iii) increase localresource mobilization and catalyze co-financing, and (iv) improve access to electricity bydisadvantaged groups.
5. Policy revision. By the late 80s, it became clear that Bank projects had delivered too littleon key OMS 3.72 objectives, especially with respect to financial and environmental sustainability.So, in January 1993, the Bank introduced a new policy which stressed the desirability of operatingthe sector on a commercial basis, the importance of energy conservation, and the requisites ofenvironmental sustainability. A notable shift, this new policy, no longer emphasizing access tobasic electricity service by the poor, identifiedfive guiding principles for Bank support:transparent regulation, commercialization and corporatization, imports of services, acommitment to reforn, and greater private investment (paras. 29-33).
I. See, for exanple, Rural Electrification in Asia: A Review of Bank Experience, June 30, 1994, Report No. 13291.
4
Review of Completed Projects
6. Since the issue of OMS 3.72 in 1978, the Bank has participated in 69 power lendingoperations in 26 SSA countries. The cohort reviewed here consists of 44 credits or loans for 41completed power projects in 22 countries-a dozen of them with two operations; and only Kenyaand Zaire having more than two. Of the total Bank financing of US$5.6 billion provided to SSAsince 1978 for infrastructure, the share for electric power was 21 percent.
7. Objectives and components. The projects emphasized physical capacity and institutionalstrengthening, and co-financing was an important feature of most. Improving access to electricservice by disadvantaged groups was attempted in only five countries. New generation-particularly hydroelectric plants-and transmission had priority over expansion of distribution.In more recent years, rehabilitation had the highest priority, particularly the rehabilitation ofdistribution. On average, about 12 percent of project cost was allocated to non-physicalcomponents-mostly for studies.
8. Performance. The outcome was rated as satisfactory for 64 percent of the cohortprojects, compared with 79 percent for all Bank power projects since 1978. For 22 projectscompleted in 1989-93, the achievement of institutional development objectives was rated assubstantial in only 27 percent of the cases (38 percent Bank wide), and the sustainability ofbenefits as likely in only 36 percent of the cases (68 percent Bank wide).
9. Economic rate of return. At completion, ERRs were recalculated for only 22 of the 41completed projects. Ten of the 22 projects had ERRs below 10 percent, compared with theaverage of 12-15 percent for all Bank power projects completed since 1978. Only three hadhigher ERRs at completion than at appraisal. With few exceptions, ERRs were calculated usinginput market prices, net of taxes and import duties, and tariffs as a proxy for the economic valueof power.
10. Institutional development and technical assistance (TA). Assistance from engineeringconsultants for project design and construction supervision was the most frequent and mostsuccessful type of TA. Management assistance was most effective where expatriates were givenexecutive authority. Training occurred in various forms, but the outcomes were poorlydocumented.
11. Compliance with covenants. Overall compliance with covenants was more often partialthan substantial. The degree of compliance with important financial covenants was especiallyweak-for the collection of accounts receivable, the approval of tariff increases, and thefinancial return on fixed assets.
12. Environment and resettlement. Power projects in SSA seldom raised criticalenvironmental and resettlement issues. The encroachment of power lines and substations onnatural resources was relatively minor. Thermal generation in several countries is based onsmall diesel-fueled sets, a relatively clean technology, and coal-fired power plants, developedonly in Botswana and Zimbabwe, use appropriate technology to abate pollution. In nine of thefifteen hydroelectric projects, the environmental impact was considered insignificant at theonset, and Bank documents do not report on the impact after completion in the others. This said,three out of five cases where resettlement raised issues-hydro power projects account for about18 percent of the 190,000 people displaced by Bank-financed projects in SSA-these were not
5
handled according to present Bank guidelines even though these projects were approved afterpublication of the guidelines in the early 80s. This confirms that the lessons learned early on inGhana-to restore resettlers' productive systems and social environment (Akosombo approved inthe 1960s)-were not subsequently applied systematically elsewhere in the Region.
Program Impact
13. Least-cost development. In line with Bank policy, Bank involvement in the SSA powersector aimed to encourage least-cost development-and it was largely successful. In theaftermath of the two oil shocks of 1973 and 1979, relatively large hydroelectric plants wereperceived as the best alternatives, but the drop in oil prices since 1986 has reduced the economicbenefits of hydroelectric plants. Those financed by the Bank in the 80s would still be goodinvestments today, but gas-based alternatives now look more attractive in Nigeria, Tanzania, andC6te d'Ivoire. Economic Sector Work (ESW) and energy assessments did not always influenceproject identification: the Bank refused to finance several projects that were uneconomical, butwith the exception of Nigeria, few options other than hydro were considered.
14. Efficiency. Project cost and time overruns, while substantial, were no worse in SSAcountries than in other regions. But, the improvement in the efficiency of technical andcommercial operations of SSA power utilities was worse and, with a few exceptions, marginal.The outcome of most projects was mixed, and the sustainability of the physical improvementsuncertain-due to a lack of effective maintenance arrangements. By the early 90s, the rate ofpower losses was higher than the Bank-wide median for two-thirds of SSA countries.Productivity, measured by the number of customers or the production per employee, was alsolow.
15. Cost recovery. With few exceptions, cost recovery in SSA countries has beeninadequate and generally below that of other regions. In the early 90s, collection of electricitybills was worse than the Bank-wide median for two-thirds of SSA countries. More than 75percent of outstanding accounts receivable was more than 90 days old, with arrears tending toincrease with the billed price of electricity. In the harsh economic environment of the 1980s,enterprises and households were hard put to pay their bills-but the bulk of the arrears werefrom the public sector.
16. The financial performance of the electric power utilities of SSA countries has generallybeen inadequate. In about 60 percent of SSA countries, rates of return (RORs) on revalued netfixed assets in operation-and debt service coverage-were worse than the Bank-wide median.On self-financing, three out of four SSA countries were worse, disappointing results given theslow asset growth in the 1980s. High to acceptable RORs (as in Malawi and Zimbabwe) reflectoperating efficiency as much as the adequacy of tariffs.
17. Access by the poor. With few exceptions-Burundi, C6te d'Ivoire, Guinea, Ghana, andNigeria among them-the provision of electricity service to low income households in SSA waspursued weakly or not at all. Where it was pursued, through distribution or rural electrificationcomponents, it did not prove sustainable. Generously subsidized tariffs, often directed to low-voltage customers, favored existing customers more than new ones. But the overriding factor,depressing incomes and demand for new connections, was the poor macroeconomicenvironment.
6
18. Resource mobilization. For power projects completed since 1978, the Bank lent US$1.2billion (36 percent of total project financing requirements), and co-financiers contributed US$1.4billion (44 percent). The utilities and the governments contributed the remaining US$0.7 billion(20 percent).
Performance Factors
19. External and countryfactors. In SSA as in other regions, international interest rates,currency movements, and terms of trade shifts hurt the power sector's performance.International fuel prices also hit the finances of utilities with predominantly thermoelectricsystems until 1986, when they began to have a positive effect. In Sahelian countries, a longperiod of drought between 1978 and 1985 depleted reservoirs and eroded the utilities' financialreserves. Beyond these outside factors, a weak human resource base and an ineffectiveinstitutional framework contributed to technical inefficiencies. Despite all this, process factorsunder Bank control were highly influential in final project outcomes, as explained below.
20. Pitching project objectives. During the 80s, broad energy considerations affected projectdesigns, increasing project complexity and making priorities more diffuse. And althoughtraditional energies such as fuelwood are very important for SSA countries-and several forestryprojects addressed them-their processing and rational use deserved more systematic attentionthan the piloting of a few components in power projects.
21. In some cases, sector objectives were subordinated to the resource transfer and otherbroad goals of adjustment operations. Conversely, adjustment loans sometimes included energysector conditions: in the 80s, a total of 23 SALs were implemented in SSA countries of whicheight, in eight countries, had such conditions. But, more often than not, adjustment o?erationsmissed opportunities to address the fiscal burden of power utilities. OED evaluations show thatSALs in the 80s were most effective in reforming the energy sector when: (i) reforms wereurgent because energy consumption per unit of GNP were high and prices were very distorted,(ii) reforms were easier because energy intensity and the share of industrial consumption werehigh, and (iii) technically sound action plans had been prepared (often under previous investmentloans). The last two conditions were seldom present in the power sectors of SSA countries.
22. Eliciting borrower ownership. The borrower's ownership of objectives was high forcapacity expansion, but not for TA and institutional and policy reforms. Least owned were:utilities autonomy, large tariff adjustments, and staff reductions or adjustments in remuneration.By and large, the Bank had little success in eliciting the ownership of stakeholders in bringingabout reform. In some cases, the Bank helped perpetuate traditional approaches-although signsexisted that bolder ones would have been acceptable. In many instances, the Bank avoidedsanctioning countries even in the face of repeated breaches of covenants, rushed to lend in ahandful of cases, and approved large loans that failed to elicit even modest policy adjustments.Policy seminars organized by ESMAP tended to dwell on principles rather than concrete actions.
23. As a result, the intellectual conviction of policymakers was not tested. Politicalcommitment, when it emerged, was less the result of Bank persuasion than of fiscal crises thatafflicted, for example, C6te d'lvoire, Guinea, Ghana, and Burundi. There is no evidence of Bank
2. World Bank Structural and Sectoral Adjustment Operations: the Second OED Overview, Report No. 10870, June 30, 1992.
7
efforts to enlist the support of potential "winners" of reform and change the incentives ofpotential losers.
24. Effectiveness of technical assistance. Again, ownership-or rather its absence-was thekey factor. Bank staff are often too remote or lack the skills to supervise TA. Other donors havegreater field presence and effectiveness, but their expertise is mostly in technology. Thechallenge is to build the role of government as regulator and policymaker through innovativeapproaches. Ghana's twinning program for management assistance was successful because ofthe good interaction between a few expatriates and a new management team of young, ambitiouspeople who were strongly motivated and eager to learn. This experience is exceptional,however.
25. Co-financing and donor coordination. The Bank has been very successful in catalyzingco-financing-less so in coordinating the role and actions of other lenders and donors. Itsinfluence has been undermined when other donors financed projects it had rejected asuneconomical. And sometimes the design of TA was not closely coordinated among donors-sothat it became diffuse, included conflicting elements, and left gaps and inconsistencies.
Implementing the New Policy Guidelines
26. Success was generally modest in establishing a transparent regulatory process, introducinga business orientation and attracting the private sector. But positive trends have recently emerged.Substantial progress has been made with the effectiveness of imported management services. Inaddition, more countries are showing concrete signs of up-front commitments in recent Bankprojects.
27. Transparent regulatoryprocess. On the basis of available evidence, separatingresponsibilities between the regulating authorities and the operating companies is deficient in thepower sectors of all SSA countries. To enhance transparency, the Bank has promotedperformance contracts ("contrat plans") between governments as owners of the power utilitiesand the management teams that operate them. These contracts have clarified objectives,stimulated dialogue between government and enterprise management, and introduced bettermanagement, accounts, and audit systems. But because contract targets are not legallyenforceable, they have not enhanced the autonomy of state enterprises or resolved their majordifficulties.
28. Commercialization andcorporatization. Many utilities in the region have corporate (orsimilar) status-and in many respects behave like commercial enterprises. But, they are notexclusively or even strongly profit-oriented, and they continue to drain public resources.Nonetheless, the prospect for meaningful progress appears to be substantial, in view of thegrowing consensus on the benefits of commercializing public utilities, and the fact that theregion may be coming out of the depression of the 80s.
29. Private sector involvement. Political and economic constraints notwithstanding,privatizing of some segments of power operations appears possible-as a few SSA countriesshow. Private operators can be expected to invest if they succeed in covering their perceivedrisk-and forthright implementation of the Bank's policy would go a long way toward reducingthe perceived risk to potential investors.
8
30. Importing services. The procurement of services from sources outside countries or eventhe region is not new-including expertise for long-term planning, project design, construction,construction supervision, tariff studies and a variety of technical advice and training. But amongimported services, management contracts have the greatest potential to make a lasting impact ona utility. Are imported services not bound to be resented by the local staff? Not necessarily, ifexperience in Cote d'Ivoire and Ghana is a guide.
31. Recent projects. Of 25 ongoing power projects in SSA, 11 were approved after the newsector policy orientation was introduced, the majority including elements of the new policy,particularly for regulatory reform and for management contracts by private operators. In manycases, those elements have been up-front and disbursement conditionalities. The FY94 AnnualReview of Portfolio Performance indicated that 87 percent of these projects were likely to meettheir stated development objectives. But even though borrower commitment to these objectivesis strong, this prognosis appears optimistic in comparison with the Bank-wide 72 percentestimate for all power projects exiting the portfolio.
32. Sector restructuring. Changes in the sector's structure are also part of the agenda promotedby the Bank. With few exceptions, power utilities in SSA countries are government-ownedintegrated monopolies responsible for the transmission and distribution of all the electricity thatthey produce. At this stage, the small size of the power systems and the weakness of theregulatory framework make it difficult to recommend their break up as was done in somecountries like the United Kingdom and Chile. But, independent power producers might be ableto compete with utilities for generation in special circumstances, so there is a need to review thebarriers to their entry in the market and the rules for pricing their output.
33. Distribution should be the prime focus of attention: it is where service delivery and costrecovery happens, and unbundling distribution from transmission and generation can open thefield for operators who do not have the technical and financial wherewithal to invest large sums.It should be accompanied by its decentralization into service areas designed to make the biddingfor the franchise attractive. And if the good record of decentralized community services forwater supply and other services is an example, this type of organization is worth experimentingwith for low-voltage power supply in areas where informal systems work better than formalones.
34. Sector information. The execution of this study reveals that the Bank has valuableknowledge (though insufficient in some areas) of the SSA power sector. This knowledge hasbeen the basis for policy choices and donors' involvement. But the principles of the new policyorientations require better information to monitor and evaluate the power sector performance andto provide a basis for decisions on private sector involvement.
Recommendations
35. The new policy orientations for power lending by the Bank makes sense for SSA-itspoverty and dependence on foreign assistance do not justify relaxing quality standards. Indeed,SSA utilities are peculiarly vulnerable to bad management, and their poor performance is onereason underlying the weak fiscal situations in many SSA economies.
9
36. The Bank and other co-financiers could exercise much more leadership in promoting thesector reforms that form the thrust of the new agenda. Until such reforms take hold, guaranteesto induce large private sector investments in the power sector are likely to be very costly.
37. Based on the foregoing findings, this review has six main recommendations.
Recommnendation 1. The Country Assistance Strategies for SSA countries shouldexamine the justification for power lending and establish its coherence with Bankassistance for broad economic adjustment and for the development and rational use ofother (especially renewable) energy resources.
Recommendation 2. Except for small operations aiming at institution building orrehabilitating facilities, power lending in SSA should be avoided in countries wheresector performance is below acceptable benchmarks in key technical and financial areasand where little of the following reform platform is under implementation at the time ofappraisal:
* The establishment of a transparent and arms-length regulatory frameworkwith legal guarantees that utilities can operate with autonomy-for example,through management or concession contracts; and
* The enforcement of regulatory principles to ensure financial discipline,adequate tariffs, and incentive-based, competitive contracting of services.
Recommendation 3. When promoting power sector restructuring and privatization inSSA, the Bank should explore setting up purchase tariffs, decentralizing distribution andunbundling it from generation and transmission, using concession contracts for privateoperators and providing guarantees for independent power producers.
Recommendation 4. The Bank should nurture SSA borrower ownership of its newsector policy principles, of institutional development programs to support the reforms,and of the delivery of technical assistance by locals and expatriates. Effectivedissemination of practice and build-up of stakeholders' support should be an integral partof project preparation.
Recommendation 5. The Bank should forge strategic alliances with other lenders anddonors to obtain a consensus on the policy objectives and criteria for their involvementin SSA countries. It should also establish partnerships in the deployment of humanresources and share responsibilities in performing those tasks that would benefit fromthe diverse field assets and competencies of donors.
Recommendation 6. In collaboration with other interested lenders and donors, the Bankshould help coordinate and institutionalize a systematic effort to gather and analyze dataon the power sector in line with the principles of the new policy.
3. Areas to monitor are technical losses, accounts receivable and rate of return on investments.
11
1. Introduction
Do Bank-financedpower projects in Sub-Saharan Africa present a special case? What is the pastrecord? Do recent trends bode wellfor the application of the new sector policy introduced in1992? As a backdrop to the study, this introduction recalls the evolution in Bank policy since the1970s and describes the challenge ofpower sector development in SSA countries. It consists notso much in overcoming energy crises as in ensuring financial viability and ensuring adequatesupply reliability and efficiency.
Study Objective
1.1 Power projects are less successful in Sub-Sahara than in other regions. Is the challengespecific to Africa? Is the introduction of a new Bank policy for the sector in 1992 relevant toSSA circumstances?
1.2 The study concentrates on low income African countries south of the Sahara and oncompleted projects initiated since the energy crises of the 70s. The study aims at:
(a) establishing the outcome of Bank financed projects, particularly the extent towhich they met policy objectives, and assessing the Bank's and Borrowers'performance;
(b) assessing the progress and prospects in implementing the principles of Bank'snew power sector policy;
(c) identifying project design and process factors which have enhanced or inhibitedproject performance;
(d) suggesting approaches to improve the impact of Bank interventions in Sub-Saharan Africa.
Study Scope and Methodology
1.3 The study assesses the outcomes of a cohort of 44 credits or loans for 41 completedprojects based on OMS 3.72 which was the prevailing Bank policy for the power sector until theend of 1992. It also considers the implications of the Bank's revised power sector policy byexamining 25 on-going credits or loans, of which 11 have been processed since 1992.
1.4 Thus, the study involved a review of 69 lending operations projects in 26 Sub-Saharancountries with a special focus on six representative countries: Burundi, Ghana, Guinea, Mali,Nigeria and Tanzania. It included visits to several countries as part of OED audit programs. Itmade use of Loan and Credits documents (SARs, MOPs, Legal documents and PCRs), audits,sector strategy papers, energy sector assessments, and ESMAP studies. It also used OED,TMWS and IEN databanks and Bank policy papers and general publications extensively, as wellas UN and IMF statistics.
12
Report Outline
1.5 To set the stage, this first chapter reviews the evolution of Bank policies in the powersector since the early 70s and the underlying challenges for SSA countries. Next, Chapter 2presents an overview of the design and performance of completed projects. Chapter 3 assessesthe impact of the portfolio with respect to the policy objectives of OMS 3.72. Chapter 4analyzes the most important factors in the performance of the portfolio as well as the Bank's andthe Borrowers'. Chapter 5 assesses the prospects for the 1992 Bank's Power Sector Policy.Finally, conclusions and recommendations are presented in Chapter 6.
The Public Utilities Paradigm of the 70s and 80s
1.6 The Bank's tradition of lending for electric power dates back to 1946 when the first loanwas made to Chile. In Africa it started in 1961 with Loan No. 279 (US$8.4 million) to Uganda.By that time, many privately owned utilities in LDCs had been nationalized and most of thepower utilities benefiting from Bank loans were State-Owned Enterprises (SOEs). SSA countriesshare that trait without exception; also, most became independent nations in the early 1960s, hada low income and mostly rural economy with low per capita electricity consumption and arelatively small and fragile power sector.
1.7 All the projects completed in SSA by end 1993 were designed under the policy stated in theBank Operations Manual Statement for public utilities projects, OMS 3.72, issued in 1978. OMS3.72 set out broad policy objectives in consonance with the public utilities paradigm prevailing atthat time; namely, the Bank was to help:
* provide power service on the basis of least-cost development programs,* strengthen the sector's institutions and improve their efficiency,* increase local resources mobilization and catalyze co-financing, and* improve access to electricity by disadvantaged population groups.
The Response to the Energy Crises
1.8 OMS 3.72 largely formalized a time honored practice, one so strong that it was largelyunaffected by the oil crises of 1973 and 1979. It was changed in 1987 to incorporate environmentalconcerns.
1.9 The Bank's reaction to the energy crisis was multifaceted. It spawned reviews of lending inoil and gas which culminated in 1984 in the policy formulation of OMS 3.82 for petroleum lendingwith its emphasis on the mobilization of private sector resources. Concern with energyconservation spilled over in the power sector. In 1980, a major study, "Energy in DevelopingCountries", was completed and for the first time a comprehensive review was undertaken, not onlyof supply but of demand issues. In 1983, the Bank issued a further study, "Energy in Transition inDeveloping Countries", which confirmed and refined the Bank's energy strategy. In the 80's, theBank increased its emphasis on management of energy demand and provision of technicalassistance. In collaboration with the UNDP, it carried out a 60-country program of energy sectorassessments which were designed to serve as a framework for investment and policy decisions bygovernments and external aid agencies. That program in turn gave birth to the Energy SectorManagement Assistance Programs (ESMAP).
13
1.10 Sub-Saharan Africa received an important share of this attention. Between 1981 and 1994,41 energy sector assessments, all of them dealing with the power sector, were carried out in 37countries. Out of 57 ESMAP studies, 21 of them were about the power sector, mostly lossreduction.
1.11 As for many LDCs, the energy crises had raised a double challenge for SSA countries: onthe one hand, to develop its domestic energy resources and substitute fuel oil, on the other hand toincrease supply and demand efficiency. It was recognized that the latter provided a more broadlyavailable avenue than the former: not all countries had cheap fuel altematives, but all of themexhibited technical losses and high fuel procurement cost especially landlocked countries whichdepended on regional transport corridors.
The 1992 Policy Revision
1.12 In the early 80's, the Bank felt that no change in power sector policy was needed; thus farand more than in other sectors, investments had been by and large successfully implemented, leastcost expansion had been an established practice, human resources were increasingly skilled; all thiscould continue to be so in a context of higher oil and coal prices; energy had become so strategic asector that govermnent, utilities and consumers alike would rise to the occasion.
1.13 By the late 80's, it was evident that some fuel oil substitution had occurred but, with fewexceptions, sector performance in LDCs had deteriorated. At the same time, it appeared that Bankprojects had delivered too little on key OMS 3.72 objectives, and on financial and environmentalsustainability (para. 5.1-5.4)
1.14 At the end of 1992, the Bank introduced a new policy which stresses the desirability ofoperating the sector on a commercial basis, energy conservation, and the requisites ofenvironmental sustainability, thus implicitly de-emphasizing the access to electricity bydisadvantaged groups.
1.15 The new policy was based on evaluation results Bank-wide, including OED.4 The 1992policy paper attributes the deterioration to three set of factors:
(a) External factors such as commodity prices, access and cost of foreign loans;(b) Inappropriate governance and national policies on trade, fiscal, monetary and energy
sector matters; and(c) Enterprise-related factors such as conflicting objectives, weak management and staff
resources.
The Core Challenge for SSA Countries
1.16 SSA countries are poor and their economic growth in the 1980s has been slow. For the 26SSA countries taken as a whole, the GDP expressed in 1994 US$ was 141 billion in 1970 and 130billion in 1993, that is US$304 per capita. GDP grew at 5.5 percent p.a. in the period 1970-80,hovering around 240 billion till 1985 before plummeting to its recent low. Nigeria dominated thepicture early on but not in later years, so part but not all of this evolution can be traced back to therise in oil prices until 1980 and their sudden drop in 1986-87.
4. The most influential OED work was Report No. 8893, Colombia: the Powcr Sector and the World Bank 1970-87, June 18, 1990.
14
1.17 Many of the factors cited above to explain the deterioration of sector performance apply tothe economies of SSA. This is particularly the case for the deterioration in the terms of trade formajor export commodities, and misguided exchange rate and fiscal policies, a weak technologicaland human resources base and the dysfunctions present in many public sectors.
1.18 SSA countries have no local manufacturing capacity for power equipment but theirenergy resources endowment is substantial though unevenly distributed: technically,hydroelectric power could provide annually some 2,500 TWh, proven coal reserves are about60,000 million ton, enough for 2700 TWh/year for 100 years; crude oil and natural gas reservesare in excess of 7,000 million ton and 5 million m3 respectively; and uranium deposits greaterthan 500,000 ton have been deemed as recoverable. However, only 0.1 percent of thehydropotential is exploited; the cost of harnessing the untapped potential is high with notableexceptions such as natural gas in Cote d'lvoire and Tanzania, hydropower in Zaire and Guinea;these exceptions call for large scale projects which often exceed the absorptive capacity of thecountry as exemplified by the hydropower schemes of Konkoure in Guinea, Manantali in Mali,and Inga in Zaire.
1.19 SSA countries consume little electricity: 31 TWh in 1971, 81 TWh in 1991, i.e. afraction of the hydroelectric potential at the site of Inga in Zaire. On a per capita basis,consumption has risen and fallen with GDP: from 134 kWh in 1971 to 247 kWh in 1980 and200 kWh in 1991. These figures are less than 580 kWh for China and 350 kWh for India butcongruent with SSA's low level of development and the predominance of its rural sector. First,in all such countries, consumption of commercial energy is very low and remains so up to percapita income of $ 1000; second, the share of electricity in commercial energy consumption doesnot exceed 25 percent except in very industrialized fossil fuel poor countries.
1.20 Power consumption has been more resilient to economic recession in Nigeria whereprices are very subsidized and in poorer countries like Tanzania and Burundi where industrialdemand, which is most GDP elastic, has the lowest share in the total. Power supply seems tohave been a constraint to growth only in Ghana and Guinea where consumption fell because offrequent and massive power outages (Figure 1.1).
1.21 Power supply should not, as a rule, lag behind or race ahead of economic growth. ManyLDCs have tried to use the sector as a leading edge of development by setting unduly low powerprices with the objective of spurring economic growth. But facilitating and subsidizing access toelectricity on any substantial scale is an extraordinarily costly proposition. This policy has provenunaffordable even in medium and high income countries. Electricity generation is very capitalintensive. In most SSA countries, unit generation costs are on the high side. Power distribution isalso expensive in cities because of the low levels of consumption, let alone in rural areas where thecost typically doubles. Evaluations of electrification prograns by OED5 and others show that theirimpact on economic growth is not significant in the absence of other development prerequisites.
1.22 The priority for the power sector of SSA countries is not to catch up with acceleratedeconomic growth, as is the case in East Asia, or repair the environmental damage caused by highenergy intensity and restructure demand, as is the case in Eastern Europe. The challenge lies
5. See Rural Electrification in Asia: A Review of Bank Experience, June 30, 1994, Report No. 13291.
15
primarily in meeting effective demand without adding to public deficits or diverting scarcepublic funds for social sectors, such as education and health. Efficiency is an essential indicatorof sector performance and also helps to maximize social welfare (i.e. the sector should provide areliable service at least cost). The imperative for financial viability is not unique to Africa:many low income countries of Latin America and South Asia face the same challenge. That it isstill prominent after years of low petroleum prices shows that it is daunting and calls for new andradical measures.
Figure 1.1: GDP per Capita vs. Electricity Consumption Per Capita
450 T
400 -- Ghana
350-
e 300
i; 250 Ivory Coast
*200 tg 150 " Guinea,,sot TNigeria100 Tanzania
, 50 - C Burundi
0 X- I l I I i I
0 200 400 600 800 1000 1200 1400GDP Per Capita, 1987 Constant US S
Tail of arrow is data for 1980 (1986 for Guinea) and tip of arrow is for 1992.
17
2. Review of Completed Projects
Since 1978, projects completed in infrastructure and energy in SSA countries have receivedabout US$5.6 billion, of which 21 percent was for electric power. They promoted verticalintegration and centralized least-cost planning expansion to capture economies of scale. Theyemphasized physical components and institutional strengthening. In recent years, rehabilitationhas had the highest priority. Technical assistance was strongly oriented towards power utilitymanagement. For the study cohort, project performance was worse than for other regions. Theoutcome was satisfactory for 64 percent of the projects. Institutional development wassubstantial onlyfor 25percent (27 percent since 1989). Results were particularly poor on costrecovery. Sustainability of benefits was likely in only 48 percent of the cases, (it has declined to36 percent since 1989).
2.1 Since the issuance of OMS 3.72 in 1978, the Bank has participated in 69 lendingoperations for power projects in 26 Sub-Saharan countries. The cohort reviewed here consists of44 credits or loans associated with 41 distinct projects completed in 22 countries. Currently, 25projects are under supervision in 19 countries, 14 of which were approved by the Board whileOMS 3.72 still prevailed (Figure 2.1 and Annexes A and B). A dozen countries had twooperations; only Guinea, Kenya and Zaire had more than two.
2.2 Completed power projects worldwide received a total of US$16.5 billion in Bankfunding since 1978. Of this, SSA received about 7 percent. Of the total Bank financing ofUS$5.6 billion provided to SSA since 1978 for infrastructure, the share of electric power was 21percent.
Table 2.1: Lending for Completed Infrastructure Projects in SS Africa Since 1978Sector Number of Projects (%/6) Loan Amount US$ million (%)
Telecommunications 15 (6) 308 (5)Transport 121 (47) 2,868 (51)Oil and Gas 38 (15) 428 (8)Water and Sanitation 42 (16) 817 (15)Power 41 (16) 1,189 (21)Total 257(100) 5,611 (100)
Objectives and Components
2.3 The review of the cohort indicates that their broad development objectives and theprinciples used to achieve them were in line with those established in OMS 3.72 (Figure 2.2).Projects emphasized provision of physical components and organizational strengthening.Mobilization of co-financing was often implicit in the financing arrangements for the projects.However, access to electric service by disadvantaged groups was attempted in only 5 countries.
2.4 A combination of eight categories of physical components were included in projectdesign: generation, transmission, distribution, rural electrification, rehabilitation, general plant,interconnection of countries, and multi-national generation (Figure 2.3). Vertical integration ofpower generation-transmission-distribution and centralized least-cost planning expansion were
18
promoted to capture economies of scale. To the same end, attempts were made at the integrationof national and multi-national markets by constructing High Voltage transmission lines andgenerating plants designed to supply multi-national markets. New generation-particularlyhydroelectric plants-and transmission had priority over expansion of distribution. In morerecent years, rehabilitation in general had the highest priority, particularly distributionrehabilitation.
Figure 2.1: Power Projects Figure 2.2: Project Objectives
II Projects under supervision (1992 policy) S institutions 36% Other, 7/o
14 Projects 41 Completed
under projects
supervision (OMS3.72)
(0MS3.72)
iamse efficiency, 190/o Provide basicifiasture, 34%
Figure 2.3: Distribution of Physical Components in the Cohort
Intercoonection ofCountries/ Multi-
General Plant national Generation Transmission13. 4% 18%
Distribution10%
Rehabilitation Gener aion35°/ Rural Electrification 18%
/ 2%
Rehabilittion Generation
Distribution OtherGeneration 638% 35% Combustion 6%
23%
Hydro
Stean 61%Transmission 10%
27%
19
Figure 2.4: Distribution of Non-Physical Components in the Cohort
Technical Assistance Training Approaches
Management On-The-Job FellowshipsOperations 32% Un-specified 3% 29%
28% 29/.
Manpower Twining TrainingEngineenng & 6% It CenterConstruction 29/.
34%
Technical TrainingAssistance Approaches:
25% 19%
Studies56%
Studies
Institutional EnvironmentDevelopment 3% Investment
21% Planning
29%
RehabilitationFinances 15%
32%
2.5 On average, a significant amount, about 12 percent of project cost, was allocated to non-physical components. The project design typically comprised a mix of studies, training andtechnical assistance components (Figure 2.4). Studies, 106 of them or about 3 per project, werethe dominant component followed by 31 instances of technical assistance and 35 training
20
components. Most of the studies were directed towards investment planning. The majority offinancial and institutional strengthening studies were focused on tariffs and sector and powerutility organization. Studies on rehabilitation and environment were few. Technical assistancewas strongly oriented towards power utility management, including some managementperformance contracts. Ten (about 30 percent) of the training components were included tomeet unidentified needs of the sector, confirming that training was not part of a well plannedtraining program.
Performance
2.6 Major Ratings. The outcome was rated as satisfactory for 64 percent of the cohort projectscompared with 79 percent for Bankwide power projects (Table 2.2). This score reflects more therelatively good achievement of projects in meeting their physical objectives than the poorefficacy on institutional development or economic efficiency. Ratings of institutionaldevelopment and sustainability do not exist for all power projects Bankwide completed in theperiod 1978-1993. But a comparison can be done for the period 1989-1993: for 22 completedprojects in SSA, the outcome of 64 percent was rated as satisfactory (74 percent Bankwide); theinstitutional development impact was rated as substantial in 27 percent of the cases (38 percentBankwide). The sustainability was rated as likely in 36 percent of the cases (68 percentBankwide). Including the earlier projects in the cohort that percentage was higher (48percent)-the downward trend of sustainability ratings is another cause of concern in SSA.
Table 2.2: Power Projects Completed (1978-93): Ratings by OEDRating Number of Projects PercentOutcome:
Satisfactory 28 64Unsatisfactory 16 36
Institutional Development Impact:Substantial 11 25Modest 21 48Negligible 11 25Not Rated 1 2
Sustainability:Likely 21 48Uncertain 18 41Unlikely 4 9Not Rated 1 2
2.7 The poor institutional development performance was given little weight in the outcomerating because in most cases, these were first or second loans to the sector and institutionalchanges require efforts over the long haul. That performance often translated into uncertain andsometimes unlikely sustainability.
2.8 Economic Rate of Return (ERR). ERRs were recaluculated for 22 of the 41 distinctprojects. The results are paltry when compared to the 12-15 percent average ERRs of all Bankpower projects completed since 1978: ten are below the minimum 10 percent of Bank
21
guidelines, and only three were greater than the appraisal estimates-one project each inTanzania, Liberia and Zimbabwe (Figure 2.5). With few exceptions, ERRs were calculatedusing input market prices, net of taxes and import duties, and tariffs as a proxy for the economicvalue of power. In eight projects the ERRs were not recalculated because of lack of data. In tentechnical assistance and one sector adjustment operations the ERRs are not applicable.
2.9 The low values of the recalculated ERRs stem from a combination of several factors, themost frequent of which are by order of decreasing importance: (i) project output valued at theprevailing electricity average price; (ii) actual volume of power sales lower than forecast; and(iii) project cost overrun. In rare cases, for example in the First Power and Water Project inMali, an attempt was made to value the economic benefits by using the estimated willingness topay for electricity service. Yet, the recalculated ERR was still lower than 10 percent becausethe cost overrun and the lower than expected electricity output offset the consumers' surplus,estimated as the difference between willingness to pay and power price.
Figure 2.5: Project Economic Rate of Return (ERR): Frequency Distribution
9 9
8 87 7 7
6 I 1SAR45 5PCR/PAR
2 2 2 2
0 ~~~~~~~~~~00-5 5-10 10 -15 15-20 20 and up
Economic Rate of Return (Percent)
2.10 Consistently low ERRs in a given country cannot be dismissed on grounds that tariffs area poor proxy of project output value. Tariffs reflect, if not the reality, at least the perception bythe regulators of the consumers' willingness to pay for electricity, and it is difficult to argue thateconomic value of electricity is high but that the willingness to pay for it is low. A low ERR is atelltale sign of either one, two, or three important facts: the regulators' perception and costrecovery policy underestimates the consumers' willingness to pay, generation and delivery costsare excessive or the project does not meet an effective demand.
Institutional Development and Technical Assistance
2.11 The purpose of technical assistance (TA) is to fill gaps in local skills. Early Bankoperations in SSA were TA projects with modest investment components (e.g. Guinea, Mali,Cote d'Ivoire). In the case of Guinea and Mali, the lack of success of investment operationsresulted in two follow-on projects which for all intents and purposes were TA projects. Theperformance varied with the type of TA: project design and construction supervision, utilitymanagement, studies and training. Management assistance essentially has taken three forms: (i)"twinning" where management and backup, sometimes including the training of counterparts are
22
provided by a foreign organization; this approach was used successfully in the formative stage ofGhana's VRA in the 1960s and Nigeria's NEPA in the 1970s, and again in the reform of Ghana'sECG in the late 1980s; (ii) the "secondment" of expatriates from other well-run utilities to fillmanagement positions, without close institutional backup or short-term obligation to trainsuccessors; such an approach was used in Botswana to fill the Chief Executive Officer's andother crucial management positions; it was also used by the Uganda Electricity Board and by theSwaziland power utility to fill some key managers positions; and (iii) the "hiring of untied",qualified expatriates as occurred in Malawi.
2.12 Project Design and Construction Supervision. This type of TA is the most common.Practically all the Bank-financed power projects in SSA have included it in one form or theother, whether or not specifically financed by the Bank. Engineering consultants usuallyprovided these services. This kind of TA was generally successful even in less developedcountries, which reflects its recognized usefulness and acceptance by country counterparts. Thecountries' preference was to retain the same engineering firm for project preparation andconstruction supervision. This worked well, particularly when the consultant comes from acountry willing to finance the services, as was the case in Nigeria, Ghana, and Guinea.
2.13 ManagementAssistance: Management assistance was most effective where theexpatriates providing the assistance were given executive authority (Ghana, C6te d'lvoire) for aperiod of time. To be accepted, such a transfer needed a strong commitment by the parentministries and the local entities' management. Wherever the expatriates were not given executivepowers, new management practices did not take root (Guinea, Tanzania).
2.14 Management assistance, especially the temporary assignment of managerial authority toexpatriates, is usually strongly resisted, particularly by the organization receiving the assistance(e.g. Guinea, Ghana, Nigeria). In most cases, opposition is unlikely to disappear untilconstraints turn into crisis (Ghana, C6te d'Ivoire, Guinea), or when it is obvious to all that a neworganization would have no chance to take off (the newly created NEPA in Nigeria in the early1970's).
2.15 Studies: Studies covered a wide variety of subjects. Predominantly, they were carriedout by foreign consultants, Ghana being an exception: VRA executed some of the studiesrequired for the Power VI Project. The impact of the studies as knowledge transfer instrumentswas mixed because of problems concerning relevance, specificity of findings and absorptivecapacity.
2.16 Studies of least cost development programs, tariffs, and human resource developmentappear in many TA programs as a part of Bank-financed power projects. In most cases they arewell justified, except where they tend to substitute for action, or where they are carried outaccording to "blueprints" which fail to address some of the most important capacity problems.Ghana, Guinea, C6te d'lvoire, Nigeria, and Uganda provide examples of these weaknesses.
2.17 Training: Training components-often funded by co-financiers are found in almost allthe power projects in SSA. Training occurs in varied forms: on the job (often used to trainmanagers); formal courses and seminars, or a combination of formal instruction and "learning-by-doing" in the country or abroad; and scholarships in the country or abroad. Suppliers alsooften provide training for the operating personnel of the equipment being supplied, a generallysuccessful activity carried out both in the country during installation and the initial operation of
23
the equipment, and abroad in the manufacturing facilities of the supplier. The building oftraining facilities and the provision of teaching materials are sometimes included in Bank-financed projects. Unfortunately, the performance of training activities is only vaguelydocumented.
Compliance with Covenants
2.18 The rating of 13 recent cases shows that compliance with covenants was more oftenpartial (9 cases) than substantial (4 cases). The degree of compliance with important financialcovenants appears to have been less satisfactory than the overall ratings imply: compliance withrate of return covenants was negligible in 4 cases out of 6; for accounts receivable, it wasnegligible in 6 out of 10; and for tariffs adjustments in 8 out of 13. Compliance with suchcovenants was not given an important weight in the overall ratings on covenants and borrowerperformance for the same reason that institutional performance was discounted in outcomeratings, that is, in most cases these loans were the first or second in the sector and improvementsin finances were considered by the evaluator as needing more time than was planned at appraisal.
2.19 Non-compliance with financial covenants was as pervasive in the distant past as it wasin the above recently rated operations: the percentage of negligible compliance is sensibly thesame for ROR and tariff covenants and worse for accounts receivable. Performance was betteron self financing and on debt service because utilities had very modest targets on both andborrowing was on very favorable terms or restructured into government equity. Table 2.3 belowsummarizes performance for the study cohort.
Table 2.3: Compliance with Financial CovenantsCovenant Substantial Partial NegligibleROR on Assets 4 3 9Accounts Receivable 1 3 9Tariff Adjustments 5 5 8Self-Financing 5 4 1Debt Service 10 2 3Asset Evaluation 4 4 4
Environment and Resettlement
2.20 Bank-financed power projects in SSA seldom raised critical environmental andresettlement issues. Many were about plant rehabilitation and had negligible or positive impacts.Even when power networks expanded, they were and still are at such an early stage ofdevelopment that the encroachment of power lines and substations on natural resources wasrelatively minor. Thermal generation in several countries was based on small diesel-fueled sets,a relatively clean technology; and coal-fired power plants, developed only in Botswana(Morupule) and Zimbabwe (expansion of Huange), use appropriate technology to abatepollution. Thus, the potential for large impacts on nature and population was greatest wheregeneration was based on hydropower.
2.21 Environment. In the 15 hydroelectric projects completed in SSA, the Bank addressedwith varying degrees of detail their impact on the biota and the agricultural land up and downstream of the plants. In nine cases, the environmental impact was considered insignificant where
24
the projects were to expand existing facilities (three cases) or were too small (six cases). Theexperience for the other six projects was mixed. In Ghana, Akosombo and its expansion wereapproved in 1962 and 1969, and Kpong was approved in 1977, long before specific Bankguidelines on environment and resettlement were published in 1980 and 1984. The PCRs andPARs emphasize resettlement issues and do not discuss physical impacts. Although theseprojects were approved in 1984, project documents do not report on the impacts of Kiambere's25 km2 reservoir in Kenya and those of Ruzizi's at the borders of Zaire, Rwanda and Burundi.For Nangbeto, the University of Benin concluded that the flora and fauna did not need specialprotection because the affected species were well represented elsewhere in Togo. For Mtera inTanzania, satisfactory health programs, a nature reserve, and a fishing program in the reservoirwere already in place since 1978 as part of the previously constructed Kidatu dam.
2.22 Resettlement. The review of environment and resettlement in Bank-financed projects inAfrica since 1981 noted that power and other sectors have similar resettlement outcomes. Hydro
6power projects account for about 18 percent of the 190,000 people displaced. Evaluation ofresettlement for the fifteen Bank-financed hydroprojects shows mixed results for SSA but noworse than for other Regions.7 8 Resettlement was an issue in five SSA's power projects, but infour cases the number of families or persons to be relocated was grossly underestimated, as wasalso the magnitude of the effort to restore their livelihood: Akosombo (about 84,000 people) andKpong (not the 5,700 people estimated at appraisal but about 7,000 people), Kiambere (not 1,000but about 6,500 people); and Ruzizi (not 30 but about 2,600 families). Although the lapses forAkosombo-the 84,000 resettlers were displaced in a last minute operation at the time ofreservoir impoundment-helped improve resettlement in the Kpong project, 80 percent of thesettlers and hosts remain dissatisfied, particularly with post-resettlement services. Similar errorsand lateness occurred later for Ruzizi (no resettlement plan, allocation of land was still pendingin 1992) and Kiambere where 4,500 people received only cash compensation for cultivable landand property valued well below market. On the other hand, 11,000 people affected by theNangbeto project were successfully resettled in Togo. In sum, for three out of five cases whereresettlement raised issues, these were not handled in compliance with present Bank guidelineseven though those projects were approved after the guidelines were published in the early 80's;more importantly the learning that took place early in Ghana to restore resettlers' productivesystems and social environment was not systematically applied elsewhere in the Region.
6. Cook, Cynthia C., Ed.: 'Involuntary Resettlement in Africa. Selected Papers from a Conference on Environment and SettlementIssues in Africa.' World Bank Technical Paper No. 227, March 1994.7. Gutman, Pablo: 'Involuntary Resettlement in Hydropower Projects." IEN Staff Report, November 1993, page 29.8. World Bank: "The Bankwide Review of Projects Involving Involuntary Resettlement 1986 - 1993'. April 1994.
25
3. Program Impact
Overall, the impact of Bank lending was substantial as regards least-cost expansion andmobilization of co-financing, but limited in terms of other OMS 3.72 objectives. After the two oilshocks, large hydroelectric plants were often the least-cost alternatives. Transmission investmentcontributed to market integration. Cost and time overruns, while substantial, were no worsethan in other regions. Labor productivity was low, andplanned service objectives and standardswere seldom met. By the early 1990's, losses were still high and so were arrears, particularlythose of the public sector. Cost recovery and utilities' contribution to investmentsfumding wereinadequate. Increased access to service by the poor did not-probably could not-take place ona substantial scale given thefinancial woes of the utilities, and the stagnation ofper capitaincomes.
3.1 The impact of completed projects is assessed below with respect to each one of thepolicy objectives of OMS 3.72.
Least-Cost Development
3.2 Most of the completed projects included generation expansion and, in the aftermath ofthe two oil shocks of 1973 and 1979, relatively large scale hydroelectric plants were often thebest alternative at the time they were initiated. These large schemes provided economies of scaleand contributed to market integration, although their very size caused financial problems whendemand growth proved considerably less than originally estimated as in the case of the Mteraproject in Tanzania.
3.3 From 1973 through 1991, the share of hydroelectric capacity increased from 42 percentto 59 percent of total capacity. It is estimated that in 1991, hydroelectric generation of some28,000 GWh helped SSA to save 8.3 million tons of fuel.
3.4 ESW and later Energy Assessments had provided good quality analyses of developmentissues and options. But, these analyses did not always influence project identification. With theexception of Nigeria, few options other than hydro were considered: coal was developed only inZimbabwe; geothermal power only in Kenya; gas could have been justified in C6te d'Ivoire andTanzania but was not pursued because neither the Bank nor the country were ready for thatoption. The Bank gave temporary support to hydro projects that were marginal such as Soubrein Cote d'Ivoire, and Manantali in Mali. It is worth noting, however, that it has refused tofinance several hydro projects that were not economical (Kompienga in Burkina Faso, Konkourein Guinea, and Bumbuna in Sierra Leone).
3.5 As may be expected, the drop in oil prices since 1986 has reduced the expected benefitsof hydroelectric plants in a few SSA countries. Gas alternatives now look more attractive inNigeria, Tanzania, and C6te d'lvoire, the latter country being most vulnerable to hydrologicaluncertainties.
3.6 Still, hydro has an edge over thermal power in other countries. The annual generatingcost of three completed hydroelectric plants (150 MW Kiambere in Kenya, 80 MW Mtera inTanzania, and the small 26.6 MW Ruzizi shared by Burundi, Rwanda and Zaire) ranges from4.6 to 7.7 US¢ per kWh. For large 40 MW diesel and 30 MW gas turbine generating plants,
26
built in land-locked places of SSA and operated with liquid fuel from crude oil at 20 US$/bbl,the generating cost ranges from 7 to 10 US¢ per kWh, or 1 I to 17 USO per kWh if crude oilprice jumped to 30 US$/bbl. These costs will be higher for smaller diesel and gas turbine plants.
3.7 Distribution expansion received a smaller but respectable share of attention. Overall,generation (with 31 components) and transmission (with 30) dominated over urban and ruraldistribution networks (with 20). Distribution often does not fare as well in other regions for tworeasons: (i) many countries have a greater problem in coping with generation requirementswhich are very capital intensive and (ii) current Bank power lending practices are not welladapted to projects which involve smaller, dispersed components, use little foreign exchange andcall for relatively less procurement through international competitive bidding.
Efficiency and Institution Strengthening
3.8 Project cost and time overruns, while substantial, were no worse in SSA countries thanin other regions. But the improvement in the efficiency of the technical and commercialoperations of SSA power utilities was, with a few exceptions, only marginal. By the early1990's, most SSA countries fared worse than the Bank-wide median on losses, accountsreceivable, self financing, and to a lesser extent on debt servicing and returns on fixed assets.Electricity prices ranged from the very low (US¢2/kWh) to the very high (US¢25/kWh) and veryseldom covered the full cost of service.
3.9 Time Overruns. Of a total of 41 completed projects, 17 experienced time overruns inexcess of 50 percent and 6 in excess of 100 percent. The 49 percent overall average time overruncompares well with the 60 percent overall average time overrun for all Bank power projectssince 19749. None of the lumpy infrastructure projects were completed on schedule, but a fewSECALs, Technical Assistance and Engineering operations were. The most frequent causes oftime overruns include: delays incurred by the Borrowers in meeting credits/loans effectivenessconditions and in providing counterpart funds; protracted procurement processes; bureaucraticdelays affecting the importation of goods; unexpected technical problems; deficient projectmanagement; and delays in carrying out studies. Less frequent causes have been the suspensionof credit/loan disbursement for Borrower non-compliance with major covenants or Governmentfailure to meet its obligations with the Bank.
3.10 Cost Overruns. Forty-one percent of projects had cost underruns, 24 percent had costsabout the same as the original estimate, and the remaining 35 percent experienced overruns aboutevenly distributed in the range of 10 percent to 50 percent (only one project having exceeded 50percent). Compare with 30 percent cost overrun experienced by all Bank power projects since1974.10 The most frequent causes of cost underruns were cancellation of project components forlack of financing (shortage of counterpart funds or failure of co-financing to materialize) andlower than estimated price of major equipment procured through international competitivebidding. The most frequent causes of cost overruns were: underestimate of base costs andcontingencies at appraisal, claims by contractors of civil works and additional engineering andadministration costs induced by time overruns. Importation of services tended to be beneficialfor engineering and construction supervision.
9. OED Report: "1992 Evaluation Results", Fig. 1.11, page 7310. OED Report: "1992 Evaluation Results", Fig. 1.12, page 74
27
3.11 Rehabilitation vs. new capacity. Rehabilitation of power plants and distribution systemswas carried out (Guinea, Ghana, Tanzania, and Zaire), but sometimes after less urgentexpansions (Tanzania, Zaire). In Ghana, VRA was particularly successful in assisting ECG torehabilitate its northern distribution network. The outcome of most projects was mixed, and thesustainability of the physical improvements uncertain (Tanzania, Guinea)-due to lack ofeffective maintenance. In other cases, the outcome was unsatisfactory as the targets were onlypartially achieved and achievements unlikely to be sustained (Zaire, Sudan). By any standards,total electricity losses are very high, ranging between 15 percent and 20 percent for mostcountries, with a few in the 30-40 percent range. The loss rate is higher than the Bank-widemedian for two-thirds SSA countries (Table 3. 1). The lack of disagregated data makes itimpossible to assess the extent of non-technical losses by cause, i.e., illegal connections, under-billed and unbilled electricity.
3.12 Labor productivity. Productivity measured by the operation and maintenance costs, thenumber of customers per employee, or the production per employee was also low(Table 3. 1). Toaddress these deficiencies, most projects chose the route of skills enhancement, but theirapproach was piecemeal and did not form part of comprehensive manpower developmentprograms embracing the entire sector. In the aggregate, the impact of the efforts deployed rangesfrom modest to negligible. As a way to circumvent the weak human resource base and to gaintime to redress this situation new options have recently been considered: management contractsin Cote d'lvoire, Ghana, Mali and Guinea; independent power generation in Cote d'lvoire.
3.13 The operating indicators presented in Table 3.1 are generally used to gauge productivity.The number of consumers and MWh per employee depend to a great extent on the general
economic conditions of individual countries and the small size of their power systems. Still, amajority of the countries under review exhibit ratios below the Bank-wide median.
Table 3.1: Operating Indicators in the Early 1990'sConsumers Per MWYh Per
Total System Losses ("l) Employee EmployeeSSA High 40 122 1308
Low 7 9 5Median 18 31 240
Bank-wide Median 15 104 602
SSA cases above Bank Median (%) 67 5 15
3.14 Service Objectives and Operating Standards. Service objectives and operating standardsare embodied in laws, decrees, regulations, policy statements, and concession, management,performance, and other contracts. They purport to define the service area, including who shouldbe served, and the quality of service. All the countries under review have a more or less elaboratebody of objectives and standards. Among the obligations of the producers, the planned qualityand reliability of service has seldom been met. In extreme cases, low quality of service causesindustry and individuals to install back up auto-generation beyond the extent that would beeconomically justified. In 1985, for instance, Guinea had a power generation shortfall of about47 percent of estimated requirements; load shedding had become a daily reality. From 1983 to1992, the private sector installed for its own use some 70 MW of power generation, and in 1993
28
produced some 109 GWh of electricity, i.e., almost as much as ENELGUI, the nationalelectricity enterprise.
3.15 TariffSetting. Technically, electricity pricing is complex. Thus, in most countries, thework of consultants has often been used as "anchor points" from which subsequent adjustmentshave been made, frequently based on a cost-of-living or other price indices. Long Run MarginalCost (LRMC) has usually featured prominently in these tariff studies. LRMCs provide goodguidelines for improving tariffs structure but must be vetted by financial projections before theycan be used as targets for setting average tariff levels. In C6te d'lvoire, capacity was abundantenough and LRMC so small that LRMC is not viable. This case is typical of many SSAcountries where growth has been slow and capacity expansion has increased debt serviceburdens.
3.16 In spite of the many pricing studies which have been carried out, information on electricsupply costs in SSA is unreliable and insufficient to allow meaningful comparisons of revenueswith cost of service. But, a comparison of average revenues across countries does provide anindication of the relative effort at cost recovery achieved. Table 3.2 and Figure 3.1 (below)portray the ranges of average revenue per kWh and compares the experience of the early 1 990swith that of the late 1970s. In some countries, prices show a decreasing trend towards levels aslow as 3-4 USO/kWh. Such prices are lower than LRMC and prices in most countries. In Asiaand Latin America, with a few exceptions (China, Nepal, Vietnam, Ecuador, Honduras andTrinidad and Tobago), the price of electricity is in the range 7-14 US p/kWh 1, while in OECDcountries 13 USO/kWh is the average tariff. 2 Only a few SSA countries collect 18 to 22USO/kWh: Benin, Senegal, Niger and Togo.
Figure 3.1: Average Revenue
5 T4-
0 -M0-2 4-6 8-10 12-14 16-18 20-24
UScents/kWh
3.17 None of the countries under review has incorporated tariff setting rules in their laws, orset up independent tariff boards. Instead, governments have formulated their tariff objectives inpolicy statements, issued mostly prior to external borrowing, including the World Bank.However, due to political expediency, such statements have often been diluted or ignored.
3.18 Another difficulty has been the lack of up-to-date financial information. In many of theSub-Saharan countries, sector entities can run out of cash long before financial statements areavailable. Tariff adjustments are frequently made on the strength of tentative financial
I1. Asia, Technical Paper Number 248: "A Survey of Asia's Energy Prices"; LAC Technical Department, Report No. 7.12. OLADE: Energy Statistics 1992
29
projections, mostly on an ad-hoc basis, too late to enable an entity to meet its obligations,requiring additional measures such as the rescheduling or conversion into equity of part or all ofthe debt of the sector.
3.19 Accounting, billing and collection. Without exception all Bank loan/credits includestandard covenants aimed at ensuring the adequacy of accounting, billing and collection.However, in SSA the results are not encouraging. In the early 1990's, collection of electricitybills was worse than the Bank-wide median for two-thirds of SSA countries (Table 3.2). Morethan 75 percent of outstanding accounts receivable was more than 90 days old: 24 percent wasbetween 90 and 150 days, and the remaining 52 percent was over 150 days. While there is aperception that collection efficiency was no better in the early 1 980s, the fragmented nature ofavailable information precludes a meaningful comparison.
3.20 Arrears tended to increase with the billed price of electricity: both were small inZimbabwe and Malawi; both are high in the CAR, Benin and Mali. However, the worstperformance belonged to Nigeria with 460 days for a low price of 3.2 cents/kWh and SierraLeone 390 days for 10.8 cents/kWh.
Table 3.2: Financial Indicators in the Early 1990'sDebt Outstanding
Self- Service AccountsFinancing Coverage Receivable
ROR on Revalued Assets (°/) Ratio (°/) (Times) (Days)SSA High 15 80 4.3 462
Low -16 -534 -1.2 60Median 4 4 1.1 131
Bank-wide Median 6 16 1.4 84
SSA cases above Bank Median (%) 41 23 39 67
3.21 Ineffective billing and collection systems, weak management, and lack of penalties orwillingness to apply them to delinquent customers are often cited as the cause of this chronicproblem. In the harsh economic environment of the 1 980s, enterprises and households were hardput to pay their electricity bills-the bulk of the arrears were from the public sector. Theproblem was exacerbated by government not allowing the power companies to withdraw servicefrom public sector entities for non payment. In this, the repeated efforts of the Bank (covenants,dialogue and other understandings) particularly in the periods preceding the approval of a newcredit, produced only short-lived results. It is not exaggerated to conclude that until governmentsregain fiscal control, and adopt a more commercial outlook, the problem of public sector arrearswill continue.
3.22 A handful of countries have started using management contracts as an expedient way tointroduce commercial discipline in the power utilities. This represents a major change in policy.The first albeit still very young experience is in Cote d'Ivoire. The private company nowrunning the power system under a management contract, has an overwhelming interest incollecting revenue from customers as it earns its fee only on amounts effectively collected. In FY92/93, its collection rate for private consumers was around 98 percent up from 63 percent in1988; due to government non-payment its overall collection rate reached only 85 percent.
30
Government arrears caused the company to retain about one-third of the money which shouldhave flowed to CAA, the government agency responsible for servicing the debt of the sector.Despite specific budgetary measures introduced earlier to ensure that the budgets of individualministries would cover their electricity consumption, their bills have continued unpaid. Similarmanagement contracts have recently been signed in Ghana, Guinea and Mali. Use of advancedtechnology, such as hard-to-tamper-meters with prepayment devices, is being considered inTanzania as a way to avoid arrears and meter reading, as well as to decrease collection costs.
3.23 Financialperformance. All loan/credits in SSA have financial covenants which promoteadequate returns on assets in operation, or pre-determined self-financing ratio. Few exceptionsexist where these objectives have been met: C6te d'lvoire, Ghana (Volta River Authority only).The financial rate of return (ROR) is of special interest when compared to the prevailingopportunity cost of capital. However, to be meaningful the ROR must be calculated on the basisof revalued assets, a condition which seldom obtains (Box 3. 1).
Box 3.1: Nigeria: Recovering Cash Costs and Depreciation
"Electricity tariffs had remained unchanged between 1979 and 1989. Averagetariffs were well below the LRMC and inadequate even to cover cash operating costs.Substantial increases in mid-1989 enabled NEPA to meet operating costs (with depreciationcalculated on historical cost) and debt service (after conversion of long-term debt intoequity and a grant equivalent to 40 percent of the costs of building a new hydroelectricpower scheme). Yet on a revalued assets basis, a further 25 percent increase would havebeen needed to break even. Without tariff increases, NEPA's deficits increased rapidly".
Source: ESMAP Report of July 1993.
3.24 An analysis of available indicators in the early 1990s points to a generally inadequatefinancial performance. In about 60 percent of SSA countries, RORs on non-revalued net fixedassets in operation-and debt service coverage-were worse than the Bank-wide median (Table3.2). On self financing three out of four SSA countries fare worse; disappointing results giventhe slow asset growth of SSA countries in the 1980s. RORs range from 2.5 percent to 7.5percent, an improvement over the 0 percent to 5 percent observed in the late 1970s. Kenya,Malawi, and Zimbabwe are exceptions with RORs around 15 percent, while Nigeria, SierraLeone, Rwanda and Guinea obtained negative returns in the range of minus 7.5 percent to minus16 percent.
3.25 High to acceptable RORs as in Zimbabwe and Malawi, which charge less than 50/kWh,reflect operating efficiency as much as adequacy of tariffs. And some countries exhibitednegative RORs even with average revenues over 100/kWh. These are mostly small systems as inSierra Leone, Guinea and the CAR.
3.26 To ensure an adequate threshold rate of return on revalued assets is a difficult goal toreach in a period of high inflation or after a major currency devaluation. To generate sufficientcash internally to finance the largest possible portion of the cost of investments can be a far from
31
satisfactory approach due to the "lumpiness" of investments and the long-term nature of powerfacilities (Box 3.2).
Box 3.2: Tanzania: Changing Revenue Covenants
Between 1978 and 1994, the revenue covenant for the electric power sector oscillatedbetween a rate of return and a cash flow. It changed three times, each time because theprevious covenant could not be met. Poor operating efficiency, inflation, devaluations, anduntimely tariff adjustments resulted in TANESCO, the electric power sector operator, rarelymeeting its revenue objectives, and when it did it was usually after a debt to equityconversion. In this case, the nature of the covenant used had no discernible impact on theutility's results.
Source: PAR on Tanzania of June 1994. Fourth Power and Power Rehabilitation Projects.
3.27 The facilities of the electric power sector are long-term assets and for this reason long-term business and financial planning are of paramount importance. The lack of integratedbusiness planning within the utilities is one of the greatest weaknesses in SSA. Investmentplanning and financing are sometimes carried out entirely independently of the operating utilitiesand of the willingness of the consumer to pay. If tariffs that are already insufficient to covercosts are too high to be increased, what is the justification for expanding capacity? Yet thishappens, while governments allow the utility to default on its debt, and to reschedule or convertdebt into equity, thus tolerating inadequate tariffs, hiding operating inefficiency, or both.
3.28 The implementation of a viable operating and financial policy would also requirereadily available foreign exchange for servicing debt, the timely acquisition of spare parts andrepair material, and investing temporary cash surpluses, a condition which has been rarelyobtained in the Sub-Sahara region.
Access by the Poor
3.29 A falling or stagnant electricity consumption per capita (ECPC) in many SSA countriescould be a sign that more small customers have been connected. Progress was made onconnections, but in most countries, it was too modest to affect ECPC. With few exceptions-Burundi, Cote d'Ivoire, Guinea, Ghana, and Nigeria among them-the provision of electricityservice to low income households was pursued weakly or not at all. Where it was pursuedthrough distribution or rural electrification components, it did not prove sustainable. In Burundi,the addition of 6,225 connections brought up to about 13,000 the total number of connections tohouseholds in Bujumbura in the period 1985-90. In Cote d'Ivoire, Bank and co-financierlending helped electrify more than 120 villages and connect about 2,700 low income householdsand their health centers, schools and community services in isolated grids and "reseauxaraignees". In Guinea, 10,000 households were added by 1986 with Bank financing. RecentBank lending is assisting Ghana in implementing the District Capital Electrifications and theSelf-Help Electrification Program: until 1994, 20 districts capital have been connected to thegrid and 240 towns/villages located within 20 km from the supply routes of the lines of thenational electrification schemes. In the latter case, the towns and villages have made materialcontribution (poles, labor) to the construction of the lines. Earlier, in the mid-seventies, Bank
32
lending assisted Nigeria in increasing residential connections by more than 600,000 in 43 citiesand 41 towns. In the 1980's the addition of 930 distribution transformers has electrified as manyvillages.
3.30 Massive connection programs such as Nigeria's became the exception in the 1980's. Butthe economic crisis depressed incomes-mostly rural incomes-and demand for newconnections. Also, cost recovery policies in SSA have generally tended to protect existingcustomers, wealthy or not, even when the need to reduce utilities' deficit was pressing.Everywhere, generously subsidized tariffs existed for most, if not all, low-voltage customers.1 3
That they benefited the landlords rather than the poor consumers is plausible given the pattern ofurban housing tenure in many SSA countries.
3.31 A household connection to the grid is expensive (e.g. $600 in Burundi, $240 in Coted'Ivoire). In a wide range of income, it is the recovery of that cost, not the price per kWh, whichconstituted the greatest hurdle to getting electricity. More consumers could afford it, and moreutilities could recover their cost if it were spread over the life of the distribution system. In thisregard, a worthwhile lesson may be borrowed from the water supply sector of C6te d'Ivoirewhere, in the early 1990s, the cost of connection for those with limited consumption wasincluded in the general water tariff, and "free" connections made. The result was dramatic: inthe space of two to three years, about 30 percent more customers had been connected.
Resource Mobilization
3.32 Since 1978, investment in Bank-financed, completed projects in SSA reached US$3.3billion of which the Bank and the co-financiers funded 36 percent and 44 percent respectively(Table 3.3). For on-going projects, commitments amount to US$2.3 billion of which the Bank'sshare is 49 percent and the co-financiers' is 46 percent. In Asia and LAC regions, the Bank hasfunded respectively 24 percent and 31 percent of Bank-financed power projects costs.
Table 3.3: Financing of Power Projects in Sub-Saharan Countries in US$ million (%)Bank (%) Co-financing (%) Country (%) Total (%)
Completed projects: 1,189 (36) 1,447 (44) 634 (20) 3,273 (100)In Supervision: 1,136 (49) 1,066 (46) 121 (5) 2,323 (100)Total 2,325 (42) 2,513 (45) 760 (14) 5,596 (100)
3.33 The above figures show a steep decrease of the countries' share from 20 percent in thepast to about 5 percent for on-going projects. However, as the countries make good theirobligations to provide financing for costs overruns, typically 20 percent of total costs, therespective shares of financing may be expected to remain essentially the same as in the past. Co-financing on the order of 45 percent of project costs was substantial and the Bank deserves someof the credit for its catalyzing presence. The countries' relatively low contribution to projectfinancing stems from their poor overall economic performance, low cost recovery, and in theweakness or absence of local capital markets.
13 . In Ghana, for instance, most of the low income residential household consume less than 200 kWh per month, and one quarter ofthem consume less than 52 kWh. Yet, the life-line block is 100 kWh per month at a flat USS0.87. Every consumer gets it.Consumption above that block is priced only 3 cents/kWh.
33
3.34 Co-financing since 1978 involved as many as 39 co-financiers comprising 28 multi/bi-lateral lending institutions, nine countries, and two private lenders (Figure 3.2). Next to theBank, AFDB contributed US$ 266 million and EIB US$ 203 million. Other co-financiers havingprovided US$ 30 million or more included CDC, CFD, KFAED, CIDA, KfW, NORAD, andSIDA.14 In terms of the number of projects in which co-financiers participated, EIB heads the listwith 14, followed by CFD with 8, KfW 6, CIDA 5, AFDB 5, CDC 3, SIDA I and KFAED 1.One can speculate that the lending institutions and countries which participated in only one ortwo operations since 1978, do not have a strong commitment to the region's development.
3.35 In the 6 selected countries the foreign debt of the power sector has decreased as apercentage of the total country debt in the period 1987 to 1991: from 13.4 percent to 10.5percent in Burundi; from 8.6 percent to 7.7 percent in Cote d'Ivoire; from 4.9 percent to 3.3percent in Guinea; from 12.6 to 9.2 percent in Mali; from 17.2 percent to 6.1 percent in Nigeria;and from 8.0 percent to 7.4 percent in Tanzania. In absolute dollar terms it has remained more orless constant in Burundi, Ghana, Guinea and Mali, but it increased by 42 percent in C6te d'Ivoireand decreased by 42 percent in Nigeria. Its composition has shifted significantly from bilateralto multilateral sources of financing: the ratio of multilateral to bilateral to other sources offinancing was 30:23:47 in 1987 and changed progressively to 48:25:27 in 1991.
Figure 3.2: Completed Projects - Contributions by Major Co-Financing Institution
300 266
250 203
200
150 86
G 100 60 46 45 34 3050
0 AFDB EIB CDC CFD KFAED CIDA KfW NORAD SIDA
Co-Financier
14. USS30 million equivalent is the average amount lent by the Bank in 41 completed projects
35
4. Performance Factors
External and countryfactors-the oil crises, macroeconomic instability and recession-didinfluence project outcome and impact. But all of these factors also existed in other regions.Process factors often viewed as defining the Bank relationships with SSA played a major part.Pitching project objectives created difficulties: the introduction of non-power componentsincreased project complexity, and the numerous adjustment operations seldom supported sectorobjectives. Borrowers did not own institutional and policy objectives, except in rare cases, whenthe situation had become dismal. No evidence was found that the Bank had a significantinfluence on nurturing ownership. Bank tolerance of inadequate compliance with covenantscompounded the problem. Except for some studies, technical assistance was generally relevant,but seldom elicited borrower commitment; where it did and consultants adapted well to localconditions, success occurred, even in the delicate area of management assistance by expatriates.Donor coordination was reasonably good on project implementation, but competition among
donors sometimes undermined the Bank's policy and complicated the design of TA.
External and Country Factors
4.1 Most factors influencing project performnance are usually grouped into three categories:(i) external factors, (ii) country and sector characteristics and (iii) process e.g. design andimplementation variables. Most external factors such as international interest rates, currencymovements and terms of trade have played a role in power sector and project performance in sofar as they influenced the cost of capital, fuel, spare parts and other imports of the sector as theydid for the economy at large. This influence was definitely negative. In many cases, localcurrency devaluations caused cost underruns in US$ terms; in local terms, they caused overrunsand increased the debt burden (Ghana, Tanzania). International fuel prices also had an influencein utilities finances for predominantly thermoelectric systems (Senegal, Sierra Leone); it wasnegative until 1985, positive afterwards.
4.2 Regional/country factors played a role as well. Indeed, the influence of fuel prices wasmasked or aggravated in many countries by fluctuations in the available hydroelectric energy; forinstance in Sahel countries, a long period of drought between 1978 and 1985 depleted reservoirsand eroded utilities' financial reserves; it also hampered economic and power demand growth.
4.3 The weakness of the human resources base in SSA countries is often cited as a problem;it may partly explain some of the technical inefficiencies encountered in project execution andutilities operations (high losses, low capacity availability). Still, the comparison with StateElectricity Boards in India, where skilled labor is not scarce but similar inefficiencies arefrequent, indicates that more factors are at play here, in particular the institutional frameworkwhich determines autonomy and incentives in management and execution alike.
4.4 All external factors and many of the country factors listed above were shared by manycountries in other regions. In particular, Latin America was rocked by macroeconomicinstability as much if not more that SSA. If there is a specific difficulty with Africa, it musttherefore play out in the Bank and the Borrower performance. This performance was rated assatisfactory in a high percentage of cases because of the good execution of physical components.The evaluation would have been less favorable with regard to policy and institutionalcomponents and covenants. With this in mind, it is appropriate to focus on process factors of
36
performance which are often seen as central when reviewing project outcome in connection withthe quality of Bank-country relationship in the Region. These are:
* pitching project objectives* obtaining borrower ownership,* optimizing Bank involvement,* effectiveness of technical assistance, and* co-financing and donor coordination.
Pitching Project Objectives
4.5 Bank-country relations in the electric power sector could be expected to be affectedprimarily by the dictates of OMS 3.72 which lists the broad objectives to be pursued by staff inlending to the sector.
4.6 Power and Infrastructure. Bank objectives as stated in OMS 3.72 were the same for allutilities: power, telephone and water/sanitation. Strictly though, the emphasis should have beendifferent e.g. access by disadvantaged populations can be justified more easily for water andsanitation than power-an expensive energy with often cheaper substitutes-or telephone. Also,achieving reasonable cost of service is often possible at a local scale for water and telephone,almost never for power. Differentiation in project identification and design should havefollowed, but the Bank displayed the same propensity to foster planning at the national level forall sectors and to overlook the cost and socioeconomic justification of underpricing electricityalmost as much as it did for water.
4.7 Power and Biomass. Objectives and components dealing with traditional energyresources started appearing in the power projects of the 1980's: This occurred in Mali where theprojects also dealt with water supply and in Burundi, Madagascar, Tanzania and Zaire wherethey included biomass and stoves components. The latter were mostly pilots and they met theirobjectives in 2 out of 3 cases. In the same period, about 19 agriculture/forestry projects featuredfuelwood plantations which were successful in 70 percent of the cases. The inclusion of biomasscomponents was not overly detrimental to the outcome of the biomass or the power components,but it tended to make priorities more diffuse and demand a broader range of skills that often wasnot available to supervision missions.
4.8 Traditional energy is essential to most of the population of SSA countries. It raises notonly a need for fuelwood plantations, but also issues of rational distribution, pricing and use.ESMAP dedicated 13 studies to household energies in 11 countries and 23 studies to non-conventional technologies in 14 countries (biomass, charcoal, stoves, peat). These studiesprovided fresh opportunities for the Bank not only to pilot a few biomass components in powerprojects but also to develop stand alone operations which could demonstrate its ability to designsmall but cost-effective loans in a non-capital intensive area of the energy sector. Yet, the studyhas not spotted any such operation among the completed projects.
4.9 Power and Macroeconomic Objectives. Power is also at the heart of the fiscal andforeign exchange imbalances of many SSA countries. In some cases, power projects becamevehicles for granting balance of payment relief and their power sector objectives were subsidiaryto broader goals. Sometimes, (Nigeria) project efficacy with respect to sector objectives suffered
37
as major sector principles were left aside, which a free standing power project might haveaddressed with better success.
4.10 In the 1980's, a total of 23 SALs were implemented in SSA countries of which 8, in 8countries, had energy sector conditions. More often than not, adjustment operations were missedopportunities to address in a sustainable manner the fiscal burden posed by power utilities. Forexample, the ESAL in C6te d'lvoire, correctly identified the problems of the sector and proposedapproaches that seemed appropriate. However, in seeking to be comprehensive, the operationemphasized too many issues, many of them in the oil sector, and therefore failed to do justice tothe dismal financial performance of EECI. In the case of Ghana, the power sector objectiveswere well coordinated with those of the macroeconomic changes, yet little progress was made oncost recovery. At other times (Burundi) the sector benefited and advances were made becausesectoral objectives were specifically addressed in separate power projects, which werecoordinated with broader adjustment operations.
4.11 OED evaluations'5 show that SALs in the 1980's were most effective in reforming theenergy sector when: (i) reforms were urgent because energy consumption per unit GNP washigh and prices were very distorted; (ii) reforms were easier because energy intensity and theshare of industrial consumption were high and (iii) technically sound action plans had beenprepared (often under previous investment loans), and all they needed was a decisive push bycore government entities (Finance Ministry). These conditions existed less for power than forpetroleum and less for SSA countries than for other adjusting ones. Indeed, it is not untilrecently that ESW and ESMAP studies have focused on the last one, that is, on preparingcountry specific reform action plans for the power sector.
4.12 Conclusion. All of the above point to the need for a carefully designed countryassistance strategy based on substantive sector work cooperatively undertaken with theBorrower. The pursuit of objectives at various levels is inevitable given the Bank agenda, buthow many can be successfully tackled in one single project? Power projects are not the bestinstruments to deal with fossil fuels or rural energy issues. Similarly, public sector managementissues should be addressed as part of institutional building interventions requiring limitedlending resources, if any. Power projects should support the broad strategy but focus on powersector objectives with up front actions designed to ensure appropriate Borrower commitment.
Eliciting Borrowers' Ownership
4.13 Country commitment is a key determinant of quality of projects at entry.'6 This reviewsuggests that ownership of objectives varied markedly with the nature of the objective and thestakeholders:
* expansion of supply was generally well owned by government and even more by theconcerned utility;
* expansion of facilities was owned more than their rehabilitation, and powergeneration more than electricity distribution;
15. World Bank Structural and Sectoral Adjustment Operations: the Second OED Overview, Report No. 10870, June 30, 1992.16. Effective Implementation: Key to Development Impact, World Bank, October 2, 1992.
38
* institutional strengthening was owned by the beneficiaries, but the degree ofownership varied with the know how to be transferred and the mode of delivery; and
* policy and institutional reforms (tariff increases, reduction of arrears, reduction ofover-staffing, deregulation) were owned less by the Government than by the powerutilities.
4.14 The Least-Owned Institutional Measures. The least owned institutional measures in thepower sector usually have been the most crucial- the achievement by the utilities of areasonable measure of autonomy, large tariff adjustments and staff reductions or adjustments inremuneration.
4.15 The resistance to increased autonomy stems in part from the capital intensive nature ofthe sector which gives tremendous clout to whoever controls the investment and financingdecisions. Also, where Governments retain the ownership of sector assets-the general case-they must be involved if only to discharge their fiduciary responsibility and keep the powersector solvent. This and the utilities' inefficiencies usually provide the excuse for governmentintervention in day-to-day operations which in turn perpetuates insolvency and inefficiency. Inthe case of multi-national undertakings such as Ruzizi hydroelectric plant, the interestedcountries have been reluctant to curtail prerogatives or to delegate them to the entity whosecontrol they share.
4.16 The measure that Governments usually resist most are electricity tariff increases. Thehuge size of the increase needed to make a dent in the sector deficits makes it impossible to do itin one installment. The need for such adjustments comes mostly from past failures to adjusttariffs regularly in small increments. A major obstacle is the general belief that electricityshould be "affordable", meaning highly subsidized for most of the customers, especially whereoperations are inefficient. There is also the practical difficulty to establish a transparent relationbetween the cost of service and the electricity rates paid due to lack of cost accounting. Finally,it is difficult to justify increased tariffs when service is of decidedly poor quality and operatingefficiency is obviously poor.
4.17 Measures involving staff reductions and changes in remuneration, though often agreedon paper are rarely owned by the entities which have to implement them. This is not only thecase because of the social repercussions in an environment with high unemployment, but alsobecause it is often politically impossible to deal with these issues in the power sector alone.
4.18 Four power projects in Ghana provide an excellent example of different degrees ofownership of objectives by the various stakeholders (Box 4.1).
4.19 Ownership of reforms. OED's work on experience with policy reforms shows that theownership stems from several factors: (i) locus of initiative, (ii) intellectual understanding ofissues by government, (iii) commitment to act and spend political capital if needed, and (iv)acceptance by major stakeholders. The Bank had little success in eliciting the ownership ofstakeholders in bringing about reform in the SSA power sector. It did not pay systematicattention to the four above factors.
39
Box 4.1: Ghana: A Case of Mixed Ownership of Power Projects
The last four power projects financed by the Bank (ECG Rehabilitation, VRANorthern Grid, ECG Power V and VRA Power VI) provide an excellent example of (i) thedifference in ownership by Government and the utilities of the objectives and strategiesunderlying the projects and (ii) the effect of coercion without detailed plan. Indeed, therehabilitation of ECG's facilities, the development by VRA of ECG's Northern Grid, and theinstitutional reform of ECG, were objectives which Ghana fully owned. Even therestructuring of ECG, which involved the temporary takeover of ECG's management by aforeign group was endorsed by Ghana. Success occurred in all three areas. In contrast, inthe financial area, where the political will to adjust tariffs was limited, where ECG'sdetermination to maintain pressure on its customers was weak, and where the Bank hadrelied on covenants without spelling out (e.g. in side-letters) the modus operandi for theachievement of the goals, the results fell short of the objectives.
4.20 The locus of initiative has to be the country, and it must be perceived as such by majorstakeholders. Timing is also important: implementation schedule which suits the country maynot agree with the Bank's view. In some cases (Guinea and Gote d'lvoire) the thrust of the actionplan seemed to be coming from the Bank, who even helped perpetuate traditional approachesalthough-signs existed that bolder ones would have been acceptable. For example, in Burundi,the Bank supported a substantial but gradual reduction in the power utility's personnel, while thecountry decided to reduce staff by 30 percent-in a single step.
4.21 Obtaining intellectual understanding mobilized a lot of the Bank's efforts in the area ofloss reduction where ESMAP was very active in hand-on seminars and 17 studies in 13countries. Policy seminars were also organized, but they tended to dwell mostly on principles.As a result, intellectual congruence with government on policy issues existed but was oftenshallow: when the actual implications of the measures envisaged became clear, ownership bythe interested entities vanished. It could have been helpful to have a more thorough discussionof issues,17 and visits to countries where similar measures were originally resisted, but ultimatelysuccessfully implemented.
4.22 Political commitment to act, when it emerged, was less the result of Bank persuasionthan inescapable facts viz. the fiscal crisis that afflicted Gote d'lvoire, Guinea, Ghana, andBurundi. The reform went faster and farther where it was conducted outside the normalframework by strongly empowered units such as the Direction Generale des Grand Travaux inCote d'lvoire.
4.23 Potential losers accept reform only if they can see some long term gains or immediatecompensation to offset their losses. Stakeholders who saw themselves as potential losers(whether true or not) were existing consumers, government leaders, and sometimes even utilitystaff; they were effective in obstructing reforms until incentives changed. The immediatewinners, utilities and long term winners (would-be consumers, consumers of other services, andtax and duty payers) were never heard. Incentives for obstructionists weakened, however, when
17. A non-African example is the 1993 Seminar on Power Sector Issues in India, which took place in Jaipur in October 1993.
40
service deterioration and fiscal deficits reached abysmal proportions. Governments were thenforced to discourage rent seeking. In any event, sources of rent were quickly drying up foreveryone. There was no case of reform going ahead before that point was reached.
4.24 No evidence was found of Bank efforts to enlist the support of potential reform"winners" and change the incentives of potential losers of reform. Recently, it did change thoseincentives in a way, when it started linking its flow of aid to performance. This relatively bluntapproach was a contributing factor in Ghana, C6te d'lvoire, and Burundi. It has not worked inNigeria where the foreign exchange generated by oil exports reduced the clout of foreign lendersand donors.
Optimizing Bank Involvement
4.25 Persistence of involvement. Bank involvement in the power sector of SSA has beendriven by the perception that the sector is very important for economic growth, is demanding interms of capital and natural resources, and presents issues which touch on many institutional,efficiency, and economic aspects affecting not only the power sector, but the economy at large.The projects analyzed show that in many instances, the Bank was keen, too keen perhaps, to giveits financial support. More often than not, it avoided sanctioning countries even in the face ofrepeated breach of covenants, to the extent of sometimes undermining its own agenda: in Zairefor instance, suspension of disbursements was avoided even when it would have been easy to dogiven the modular design of the project which was about rehabilitation. In Guinea, the secondproject came too soon to extract government commitment or incorporate the lessons of the firstone.
4.26 Larger loans did not buy a better performance on policy and institutional development.For instance, five loans of about US$ 100 million were made but sustainability is likely for onlytwo of them and institutional impact was substantial for only one. Policy adjustments never metagreed targets both in Nigeria and Cote d'Ivoire where initial conditions were dismal and inZimbabwe and Kenya where a surge in capacity expansion made it urgent to improve costrecovery.
4.27 The desire to maintain a presence in the sector, to ensure a positive aid flow, or not toundermine broader scale adjustment efforts supported by the Bank may have been at the root ofthe Bank's decision not to invoke remedies or to rush a second loan. Arguments also exist infavor of making big loans: they boost the aid flow, leverage the uncompressible cost of Bankstaff involvement and can be turned into a visible show of donor coordination when several co-financiers are involved. None of these reasons is compelling if the decision for massive financialassistance does not support the ultimate objective for getting involved in a particular sectorwhich is to achieve progress on specific developmental issues. This is a test of coherence andcredibility. The power sector can seldom claim to be part of a core program of Bank assistanceand as shown by the examples of Egypt and later Colombia, Brazil and India, the Bank can verywell scale down its assistance in the power sector when reform commitment by the country hasbeen unmistakably absent. Such action can yield results, especially if it is combined with a wellarticulated argumentation drawing on the Bank's global experience.
4.28 Intensity of Involvement. Once the Bank decides to lend to a country, a substantialamount of Bank staff involvement is inevitable: it has to make sure that policies and proceduresare correctly understood and agree with Government on objectives and strategies.
41
Implementation also entails a substantive use of staff resources. Bank involvement in projectpreparation and implementation is always a balancing act between too little and too much. Toolittle involvement leads to projects that cannot be processed or to operations whose value addedis unsatisfactory; in the SSA context, this usually means a worthless project because of the loweraverage level of development, and because most of the projects under review (except perhapsthose for Nigeria and Ghana) were among the first Bank operations in the sector. Too muchinvolvement leads to a confusion of roles and responsibilities with the borrower. This reduces itsownership and increase the risk that the Bank's approach will not be fully understood, and itsintervention will be resented as overbearing. The PMTF reported that excessive Bankinvolvement was one of the borrowers' most frequent complaints and greatest source of lowperformance.
4.29 Was Bank involvement in SSA countries been more intrusive than in other regions? Inthe power sector such a perception could arise from the large number of studies and consultantsinvolved in Bank projects but not from the intensity of involvement by Bank staff which is inline with Bank averages: 180 staff weeks per project, that is less than 10 percent higher than theBank-wide average for power projects; the 90 staff weeks spent from identification to approvalwere 18 percent below average, and the 90 staff weeks spent on supervision were 27 percentabove; the average number of supervision missions per project was about 1.8 per year.
4.30 Evaluators have never rated supervision intensity as excessive. On the contrary, PCRs ofunsuccessful projects in SSA often conclude that Bank supervision was insufficient although,e.g. in the unsuccessful projects in Guinea, an average of two missions per year had visited thecountry. It is unlikely that the Bank can, on a sustained basis, field more than two supervisionmissions per project per year. It is unlikely also that a more intense Bank supervision could havereversed the unsatisfactory outcome of some projects which failed mainly because of insufficientquality at entry. It is well to remember that no amount of supervision can make up for thecountries' lack of commitment to bring about change. Evaluations of portfolio managementsuggest that the Bank should re-structure or cancel projects earlier and more often and re-focusrather than increase supervision. In particular, concentration of supervision in the first third ofthe implementation period has been shown to be strongly correlated with success.
4.31 Identification and appraisal often requires that the Bank exercise strong judgmentsabout national priorities and project quality. Although necessary and in the interest of thecountries, these are rarely appreciated at the time they are presented and seen as intrusive. Suchcases occurred but, seemingly not more than in other regions; examples include: the rejection ofthe Kompienga hydroelectric project in Burkina-Faso, of Konkoure in Guinea and of Bumbunain Sierra Leone; and the downsizing of Ruzizi in Burundi, Rwanda and Zaire. By contrast, Bankinterventions were generally welcome when they resulted in mobilizing co-financing.
4.32 Duringpreparation, the potential for a confusion of roles is greatest, and the Bank wasoften engaged in more discretionary activities, but again, it seems not more than for otherregions. The two examples in Box 4.2, one exhibiting success, the other failure, illustrate therange of such activities and their outcome, often determined more by country factors than byBank performance: in Nigeria, they failed even though they were traditional in nature while inGhana they succeeded in spite of being delicate and less conventional.
42
Box 4.2: Bank Involvement in Project Preparation
In connection with the Ghana Kpong project, the Bank played an advisory role in the re-negotiation of the contract between the Volta River Authority (VRA), a generating company, andits main client, VALCO, a private aluminum smelter. It had played a major role in theagreement of the early 1960's. The Bank also contributed substantially to the formulation of thePower System Rehabilitation project, which introduced top-to-bottom changes at the ElectricityCorporation of Ghana (ECG), including temporary management of the utility by expatriateconsultants. The Bank was also heavily involved in the unconventional assignment of VRA toexpand the national transmission and distribution grid in the northern part of the country and tooperate the regional system. This was done on the strength of VRA's institutional capability, arecognition that proved fully justified.
In Nigeria, the Bank was in a more traditional role when it tried to take advantage ofsuccessive political changes to help improve planning and management practices in the powersector. Thus, in the early 1970s, during the appraisal of the Power IV project, Bank staffprepared the first investment program and financial statements for the then recently createdNational Electric Power Authority (NEPA); in the late 1970s, in the preparation of Power V andPower VI, it tried to induce the Nigerians to balance their investment program, which over-emphasized generation; in the mid-1980s, while preparing Power VII project, the Bank providedmodels and terms of reference for studies to underpin approaches largely new in Nigeria(twinning, contrat program, pilot developments, etc.). In each instance, the Bank's interventionswere not unlike those made with success in many countries. In these cases, they wereunwelcome and, though it was reflected in project conditions, failed to produce lastingimprovements.
Effectiveness of Technical Assistance
4.33 Setting up a TA program and modalities. In the less developed countries, the variousdevelopment agencies, especially the Bank, play a major role in shaping TA programs. In theSSA context, with its relatively simple sectors and the generally obvious needs for TA, this taskdid not receive the analytical scrutiny it deserved. Experience shows that TA success is elusive,except for TA associated with design and implementation of physical project components, whichis usually implemented in accordance with the spirit and the letter of the donors' rules.
4.34 In capacity building areas, the performance of TA Programs has been affected by thedegree of ownership, the relevance of the TA component, the individual sectors' absorptivecapacity and the knowledge of (and sensitivity to) local conditions possessed by expatriateconsultants.
4.35 Ownership. The success of TA programs depends a lot on the degree of ownership bythose concerned. Countries generally assign low priority to TA and TA programs in contrast tothe urgency they give to developing infrastructure. This bias and the lack of agreement whichmay ensue between the Bank and the country are frequently resolved by postponing actionprograms unti'l after studies have been carried out. In this respect, the reform of Ghana's ECG inthe late 80s is an exceptional success story: the twinning program for management assistance
43
succeeded because of the good interaction between a few expatriates and an essentially newmanagement team of young, ambitious people who were strongly motivated and eager to learn.(Box 4.3)
Box 4.3: Good Practice on Twinning with Expatriates
A twinning arrangement between Ontario Hydro of Canada and Volta River Authority (VRA) ofGhana had made up for the lack of technical and management capacity of VRA during its earlystages of development. A team of expatriate managed the operations of VRA and trained thelocals who progressively took over. VRA has continued to maintain a cordial and closerelationship with Ontario Hydro for more than 25 hears and now gets from it the benefit of bothspecialist expertise and training facilities, often on short notice.
In a more recent twinning arrangement between the Electricity Supply Board of Ireland (ESB)and the Electricity Corporation of Ghana (ECG), ESB provided managers in a number of keypositions (Managing Director, Director of Operations, Project Manager, Co-Principal of theTraining Center), who helped ECG to prepare a Corporate Plan for the period 1989-1992 to berolled forward, and staff training. Their line responsibilities were well accepted and relationswere cordial. Success was achieved in the areas of personnel management (creation ofincentives, reductions of staff, staff development) and investment program, not in the reductionof energy losses. The expatriates were replaced by locals, after three years, as the competenciesof the latter were developed.
4.36 By contrast, in Nigeria after 1975, ownership was consistently lacking, mostly because thesupervising agency was itself weak and existing managers retained their old positions; theoutcome was an unmitigated failure. Failure also occurred in Guinea in part because expatriateswere limited to a weak advisory role to unresponsive local managers.
4.37 Relevance. Least-cost sector development and other studies often lack relevance in thatthey fail to take account of the stage of development or the level of performance of the sector. InGuinea, Mali and Nigeria through the late 80s, is was neither realistic nor meaningful to call fora traditional least cost development study assuming that demand depends essentially onexogenous factors and foreseeing coverage of such demand by a least cost investment sequence.Power sector development in these countries was rather obvious and largely determined by theirfinancial and physical implementation capacity. A simpler study would have made more sense:it could have provided an outline of short and medium term development, which could have beenthe basis for Bank-financed projects. The Bank recognized this in Nigeria in the early 1970s,when it carried out a rough analysis of a likely program on its own, and included an in-depthstudy into the project. Similarly, the Bank required sophisticated tariff studies in Guinea in theearly 1980s and Uganda in the late 1980s, where available data were of poor quality and wherereasonable estimates of tariff level and structure could have been determined at much lessexpense.
4.38 Capacity for Absorbing TechnicalAssistance: Some SSA countries face the inadequacyof basic schooling, especially for lower echelons. This not only affects the starting point but alsothe pace of carrying out a TA program. Hence, many actions contemplated in TA programs are
44
never carried out, and their planned outcome is only partially or never achieved. When inaddition to having limited capacity, a sector has embarked on a relatively large infrastructuredevelopment program, the more capable staff are very likely to be mobilized to ensure thesuccess of that program as a matter of priority. Imparting the very first knowledge is always themost difficult. Absorptive capacity should increase over time because there is synergy inlearning. This is evidenced in SSA countries where some sectors, notably the power sector inGhana, have developed a remarkable absorptive capacity.
4.39 Training needs in SSA are tremendous due to the pressure to employ as many localpeople as possible, in spite of their inadequate preparation to fill positions in a technologicallysophisticated sector such as power. One of the difficulties often faced, is the inadequacy of basicschooling, especially for lower echelon jobs (mechanics, electricians, linesmen, accountants,administrators, etc.). The question is how to overcome this shortcoming. One option could be toinclude some elements of basic schooling in the training programs of the utility to compensatefor what the education sector was unable to provide. A case in point is Tanzania, whereTANESCO, the electricity utility, provided a successful literacy program, which resulted inevery employee being able to read and write.
4.40 Another has been the lack of enlightened management and business organization capableto justly value and use fresh skills. Lastly, utilities lose trained staff mostly to the private sectormainly because of inadequate salary levels, (although the cost of training is not wasted since itpartially benefits other sectors of the economy).
4.41 Performance of Consultants. A high incidence of dissatisfaction is observed among thegovernments and sector entities with regard to the performance of the consultants employed insupport of sector efficiency and development (Burundi, Guinea, and Mali). Large consultingfirms appear to have difficulties to interface with utilities of modest size. They are said to taketoo much for granted in matters local officials feel should be discussed and explained. Selectioncriteria for such consultants should include their proven ability to work with small organizations.The most experienced consultants are not necessarily the most desirable in this environmentsince they tend to advocate sophisticated solutions and to be paternalistic. Conversely, lack orscarcity of well prepared Government and sector counterparts limits the quality of theconsultants' performance. Further, there is the never easily solved question of compatibilitybetween trainer and trainees, especially when they have different cultural backgrounds. Finally,the local managers of small entities such as exist in Burundi, Mali, and Swaziland, are oftenoverly suspicious of the motives of expatriates, especially when they belong to powerfulworldwide organizations.
4.42 Limitations of Bank Instruments. Bank staff is often too remote or lacks the skills tosupervise and help implement TA. Field offices as currently constituted cannot make up forthese shortcomings in most areas. On the other hand, donors have greater field presence andeffectiveness such as GTZ, DEH etc., but their areas of expertise are mostly in technology. TheBank has not used them enough in the past in situations where they could have beenadvantageously substituted for its own staff (Burundi), and has not been effective in forgingpartnerships with entities able to dispense experienced personnel to meet the challenge of
45
building the role of the Government as regulator and policymaker through innovativeapproaches.
4.43 Conclusions. The following conclusions are particularly, if not exclusively, valid forSSA countries:
* technical assistance need to be embedded in a program to ensure that it: (i) pursuesworthwhile objectives, (ii) contributes to synergy, (iii) is within the implementationcapability of the entities concerned, (iv) fits within the context of the local culture,and (v) is properly timed;
* management assistance involving delegation of executive power to expatriateconsultants has the best chance to succeed when (i) the overseeing authority in thecountry is fully committed to the TA program and prepared to enforce theirexecution, (ii) the expatriates have the ability to adapt their work style to the cultureof modest power utilities, and (iii) a large part of the assisted local managementteam is new in the positions which the consultant is to help them assume efficiently;
* training programs should be designed as components of a comprehensive programfor the sector. When urgent measures need to be taken without the benefit of havingsuch a program, it is important to ensure that they (i) address pressing needs, (ii) arecoordinated with other measures (e.g. those funded by co-financiers), and (iii) arenot in jeopardy because of a lack of necessary parallel or supportive action;
* training programs must be seen against the background of the general schooling ofthe trainees: some skills which usually are imparted in school may have to be taughtor upgraded in the context of training in the power sector;
* the Bank should avoid including studies as a proxy for concrete action when the typeof action needed is well known (e.g. tariff adjustments, personnel cuts, changes insalary structure etc.) but the quantitative adjustments required are uncertain. With orwithout study, it is always possible to define steps, albeit falling short of theestimated ultimate goal, that could be included as a pre-requisite for lending; and
* as a corollary, studies and the implementation of their recommendations need to betaken seriously, or left out altogether as project components.
Co-Financing and Donor Coordination
4.44 The Bank has been very successful in catalyzing co-financing-less so in coordinatingthe role and actions of other lenders and donors. The extent of its involvement in such acoordination depends essentially on its relations with and the capability and willingness of theGovernment and other interested entities to play their role in a way which is compatible with theBank's approach. Even in straight co-financing, the Bank had virtually no role to play in Nigeriawhere the country's revenues from petroleum in the 1970s and early 1980s precluded the needfor co-financing coordination; in Cote d'Ivoire because of other major participants in the sector,the Bank played only a minor role.
46
4.45 Donor competition has many advantages for borrowers, but it can be a mixed blessing,particularly for smaller countries. Because of the pressure on the developing agencies to helpthese countries in priority, plenty of money was in principle available to develop their powersectors. The kind of competition which developed for project identification among would-belenders sometimes became detrimental. The Bank influence has been undermined when otherdonors financed projects it had rejected as uneconomical (e.g. Konpienga in Burkina Faso); inBurundi, it led to delays in the construction of the Ruzizi II Regional hydro project and to thepremature construction of the Rwegura hydro plant, a project which was not the least-costsolution. In Uganda, in the late 1980's and early 1990's, the investment program was mainlydriven by the availability of funds and donors' preferences. Lately however, convergence in thepolicy agenda of the most important donors has increased.
4.46 Sometimes the design of TA was not closely coordinated among donors-so that itbecame diffuse, included conflicting elements, and resulted in gaps and inconsistencies. Forinstance, in Uganda in the late 1980's and early 1990's, the TA program was a superimposition ofessentially donor driven which taxed the country's absorptive capacity. In the same period, inRwanda, the situation was similar. In C6te d'lvoire, inconsistencies in the TA program resultedfrom a lack of clarity of sector objectives and strategies, and the differences in theirinterpretation by the country and the main donors. Here, the Bank was hampered in its efforts bythe fact that its participation was that of a minor donor and its relations with the main actor in thecountry were strained.
4.47 Coordination during implementation seems to have gone reasonably well. There wereonly two recently reported cases of unmet expectations. In the first case (Tanzania-Power IV),one of the co-lenders withdrew its contribution because the civil works contract was not awardedto a contractor of its country; another co-financier reneged on its pledge to finance TA; and twoof the co-financiers did not disburse funds because of the country's arrears to them. In the othercase (Guinea - Second Power Engineering Credit), the Bank organized and chaired donormeetings, but the Bank-financed team which was to assist in the coordination of donor financedrehabilitation did not arrive in the field until very late.
4.48 Conclusions. The Bank needs the credibility provided by a sustained and reasonablysuccessful sector involvement to effectively promote its sector improvement agenda. Anunderstanding and congruence of views among co-lenders on broad sector issues and how toaddress them are necessary ingredients for successful coordination.
4.49 Most power sector entities in the least developed SSA countries are ill equipped toprepare and implement complex financing arrangements involving several co-lenders. There is aneed for streamlining. The Bank could take the lead to introduce new approaches with little riskif operations are small as they often are in SSA.
4.50 The Bank should forge a strategic partnership in the sector and arrive at a division oflabor with co-lenders. In cases where a nearly permanent presence is needed and where the Bankmight wish to keep oversight over a particular project component, it might consider delegatingsupervisory functions to agencies that permanently have specialized staff in the country andwhich are often more experienced than the Bank in field operations.
47
5. Implementing the New Policy
The new policy stresses the commercial nature of power supply, an elusive goal in Africa TheBank is optimistic about the performance of on-going projects. However, the picture is mixed.SSA countries still lack a transparent and arms-length regulatory process, and evencorporatized utilities lack business orientation and autonomy. "Contrat plans" have clarif edutilities' objectives but do not provide a legal basis to curb government interference. In onecountry, progress has come recently from management contracts for utility operations.Importation of services has been key to this contract and to the success of twinningarrangements for management assistance in afew other cases. Most recent projects featureelements of the new policy with up front conditionality on regulation and management contracts.They signal that more countries are committed to reform. However, private investment has notoccurred yet on a substantial scale.
The New Policy
5.1 The fourth objective of OMS 3.72 - Improving Access by the Poor - was arguably theleast feasible objective of completed projects for SSA countries, given their financial and fiscalwoes, their low per capita income and their stage of economic development. This is true also inthe foreseeable future. The Bank's latest power sector policy was approved in October 1992 andpublished in January 1993. While the new policy maintained the implicit principles of economyand efficiency of OMS 3.72, it signaled a notable shift in emphasis. It also reaffirmed theimportance of Bank policies on resettlement and environmental sustainability, but abandonedaccess to a minimal electricity service as a welfare objective.
5.2 The 1994 World Development Report (WDR) argued that, given its current dominance,the public sector will continue to have prime responsibility in most countries for the provision ofinfrastructure services. This is certainly the case of SSA countries. Improving the effectivenessof the public sector infrastructure providers is thus critical, and WDR proposes instruments akinto those found in the new policy to reinforce the commercial operation of the public sector.
5.3 The new policy identifiesfive guiding principles for Bank support of power sectorrestructuring programs. These are: transparent regulation, commercialization andcorporatization, importation of services, private investment and commitment lending. Theseprinciples make up the menu from which "committed" countries are expected to choose toresolve their specific problems.
5.4 Concomitantly, a new policy was introduced to emphasize the need for demandmanagement and energy conservation objective in several sectors, one of which was electricpower. This Chapter reviews recent project trends and examines the prospects for a sample ofsix SSA countries to internalize the guiding principles of the new policies.
Trend in Recent Projects
5.5 Of 25 on-going projects in SSA, II were approved after the new sector policyorientation was introduced (Annex B). These projects are decidedly stressing rehabilitation (34components) and distribution expansion (11) over expansion of generation and transmission
18. The World Bank's Role in the Electric Power Sector, January, 1993.
48
(20). The number of studies (48) is decreasing from three to two per project, mostly becauserehabilitation and pricing all but disappeared as a topic. On average, one out of two projectsincludes elements of the new policy particularly for sector structuring and regulating,management contracts by private operators, and manpower development. Rehabilitationsgenerally involve greater energy efficiency; the projects do not include direct demandmanagement component and rely as in the past on tariff reform.
5.6 The performance of these projects has been rated as high by the Region. In the FY94Annual Review of Portfolio Performance (ARPP), indicated that 87 percent of them were likelyto meet their stated development objectives. This percentage is optimistic in comparison tomany others: that in FY93 (74 percent), the Bank-wide rating for power (84 percent), the Regionaverage for all sectors (81 percent), the Bank-wide rating for power projects exiting the portfolio(72 percent) or the historical average at completion. In the past, supervision ratings wereconsistently more optimistic than completion ones, the size of that disconnect was about 5percentage points at the last supervision mission.
5.7 A joint OPR and OED study'9 reviewed the quality of economic analysis of projectsapproved in 1993. This evaluation suggests that Bank-wide power project appraisal still suffersfrom many of the weaknesses diagnosed by the PMTF: lack of risk analysis, unrealisticassumptions about tariff increases, weak integration of environmental impact assessments intoproject design and lack of analysis of social impacts. Two projects in SSA-Tanzania, Power VIand Ghana, National Electrification-were in the sample of 10 power projects reviewed by thatstudy. They exhibited some of these flaws but were cited as good practice examples in twodifferent areas: analysis of fiscal impact and cost recovery in Tanzania and a very goodquantitative sensitivity analysis in Ghana.
5.8 Also, quality at entry depends a great deal on the relevance to the five principles of thenew policy and the prospects that these principles can be adopted by SSA countries. Themeaning of project success would be diminished if project objectives were only marginallyrelevant to these principles or too modest to make an impact. The chapter reviews below howmuch progress has been made thus far on each of these principles and in particular how far hasthe Bank gone in applying its principle of commitment lending.
Transparent Regulatory Process
5.9 Regardless of ownership of power sector assets, Government must set objectives,articulate overall policies, and coordinate sector development. It must also establish thelegislative and legal framework to protect the interests of the various stakeholders. Aspects to becovered include inter alia investment programs, pricing, access to service, reliability of service,energy conservation, environmental issues, and disclosure requirements.
5.10 The purpose of transparency is to enable the greatest possible degree of accountabilitythus prompting and enlightening public debate. Accountability means the need for and the factof answering for an assigned responsibility, an activity best performed when responsibilities areclearly defined, and information on the area of responsibility is readily available.
19. A Review of Quality of Economic Analysis in Staff Appraisal Reports for Projects Approved in 1993, Draft Report. May 5,1995.
49
5.11 The regulatory process should be arms length and ensure an unequivocal separation ofresponsibilities between the government (responsible for policy formulation, compliance reviewand enforcement) and the entities responsible for power supply. An autonomous regulatoryagency should also be instituted to enforce regulations and settle disputes.
5.12 On the basis of available evidence, separating of responsibilities between the regulatingauthorities and the operating companies is deficient in the power sectors of all of SSA countries.Government financing and ownership create budgetary and stewardship responsibilities whichrequire some government involvement, not all of which can be delegated. However, in thecountries under review, this involvement has gone way beyond discharging statutoryresponsibilities.
5.13 One must look beyond the text of the enabling legislation or the by-laws of the entitieswhich generally gives the utilities the theoretical authority to manage their sectors. In practice,the chairperson and often all of the directors of the state-owned utilities are appointed bygovernments, a cozy arrangement which makes it easy for the parent ministries to ignore theboard's authority and make decisions for the utilities (Box 5. 1).
Box 5.1: Burundi: Separation of Responsibilities on Paper
In Burundi, the power utility is REGIDESO. It is a Government-owned enterpriseunder the Ministry of Energy and Mines who has overall responsibility for policyformulation. The Board of REGIDESO is composed of Government appointees andrepresentatives from the Chamber of Commerce and the Municipality of Bumbura, thecapital city. REGIDESO's board is (in principle) responsible for all necessary decisions tomeet the company's objectives, set administrative and financial rules, regulate personnel,propose tariff changes, and prepare annual budgets. Expansions are financed in part bymulti- or bilateral agencies, with the Government making the final investment decisions.The written prerogatives of the Board notwithstanding, the Government continues to financepart of REGIDESO's expansions and to guarantee the remainder of the financingrequirements. The introduction of a five-year performance contract between theGovernment and REGIDESO in 1989 has added a new focus on improving sectorperformance, but this has not changed the locus of major decisions.
Source: World Bank - SAR, Power Transmission and Distribution Project, April 1985.- SAR, Energy Sector Rehabilitation Project, February, 1991.
5.14 Performance Contracts (Contrat Plans). In order to enhance transparency andperformance of the public enterprises, the Bank has promoted the use of performance contractsby its borrowers. These define, in more or less detail, the relationship between Governments andthe management of their public enterprises. A Bank review of some of these contracts20
concludes that: (i) their major weakness is that they are not legally enforceable, and oftenGovernments have not carried out their part of the understandings; (ii) the hope that performancecontacts would enhance the autonomy of Government-owned enterprises was not realized; and(iii) they have not proven particularly effective mechanisms to help enterprises with serious
20. John Nellis, Contrat plans and their role in the perfonnance improvements of public enterprises, February, 1989.
50
difficulties. The experience with performance contracts in Burundi, Ghana, and Senegal isrevealing: the Government still acts unilaterally and violates agreements. However, theproblems with contrat plans stem more from their implementation than from conceptualweaknesses.
5.15 In spite of their limitations (they are not legally enforceable), contrat plans have clarifiedthe objectives of many state-owned enterprises, stimulated a dialogue between Government andenterprise management, introduced better management information systems, accounts and auditssystems, and physical and financial performance indicators.
5.16 It is easy to imagine the power which can accrue from the ability to make majorinvestment and procurement decisions in a sector as capital intensive as the electric powersector. For this reason, in countries with long traditions of governments making those decisions,it may prove very difficult to change the de-facto allocation of responsibilities to empower theutilities with the desired authority.
5.17 Another reason for government interference, although possibly more amenable tochange, consists in the lack of training of the ministries in policy formulation, control, andenforcement. Yet another reason often mentioned for government involvement in sectoroperations is the opportunity for nepotism and patronage or simply to find occupation for newgraduates. The extent to which this exists is not easily ascertained.
5.18 Disclosure Requirements: Without information, markets cannot function properly or atall. Without continuous, reliable information, there can be no transparency, and hence little ifany accountability. One of the characteristics of many of the countries of the region is theinadequacy of accounting systems and the staff that operate them, and the substantial delay inclosing the books. The audited financial statements of the Tanzania Electric Supply Companyhave been issued with such delays as to be meaningless as feedback and control mechanisms.The Volta River Authority of Ghana has done well (Box 5.2).
Box 5.2: Good and Bad Practices on Financial Reporting
In Ghana, "The accounting and budgeting systems used by VRA are generallysatisfactory and provide timely and reliable financial information. The reorganization of theCorporate planning and finance function in 1990-91 has been completed and the key managerialpositions filled. Improvements in financial management and information systems, corporateplanning and reporting have been made following the full computerization of VRA's accounts.Other improvements are expected as a result of the full integration of financial management".
In Tanzania, TANESCO has been unable to meet its reporting deadlines. In FY80 andFY8 1, audited financial statements were produced one year late. Then again, audited accountsfor 1983 were not completed until July 1985; the 1990 accounts were submitted only in February1993, and the 1991 accounts in June 1993.
Source: Bank SARs-for Ghana-Thermal Power Project, January, 1994, and for Tanzania-Fourth and Sixth Power Projects and Power Rehabilitation Project.
51
Commercialization and Corporatization
5.19 The new policy states: "For power enterprises to operate on commercial principles, theymust be treated (and behave) like commercial enterprises. They should pay interest and taxes;earn commercially competitive rates of return on equity capital; and have the autonomy tomanage their own budgets, borrowing, procurement, salaries, and conditions pertaining to staff".
5.20 Experience in the region and elsewhere shows that while many wholly-ownedgovernment corporations, for example airlines, railways, etc. behave commercially in manyrespects, they are not exclusively or not strongly profit-oriented, and continue to drain publicresources. In other words, corporatizing publicly-owned enterprises is no assurance they will beoperated as business enterprises. The utility in Kenya is partly privately owned and yet it wasinstructed by government not to comply with its lease agreement for the use of the largeKiambere dam, thus causing the financial insolvency of its owner.
5.21 Many of the utilities of the region have a corporate or similar status-for example theTanzania Electric Supply Company-in many respects behave like commercial enterprises.Unfortunately, as pointed out above, many of the ingredients of business orientation are missing:investment decisions are the purview of government; tariffs are approved on an ad-hoc basis;service cannot be systematically withdrawn for non-payment; personnel policy is shared betweengovernment and the utilities; foreign exchange is difficult to obtain for current needs; and muchimprovement is needed in operating performance, financial management, and reporting.
5.22 The transformation of most utilities has been underway for a long time, with remarkableadvances and equally striking setbacks. In this, the general state of the world and local economyhave played a very major role. Nevertheless, the prospect for meaningful progress appears to besubstantial, especially in view of the growing consensus on the benefits of commercializing thepublic utilities, and the fact that the region may be coming out of the depression of the 80s.Once the general liquidity problems have been resolved, governments should be able to clear uparrears. With increased prosperity, regular tariff adjustments will become easier to entertain, andmany other aspects of commercialization will be seen as natural, and will be more readilyaccepted.
Importing Services
5.23 The procurement of services from sources outside countries or even the region is notnew-including expertise for long-term planning, project design, construction, constructionsupervision, tariff studies and a variety of technical advice and training. In the majority ofcountries, the importation of services, including the training of nationals in foreign countries orthe use of foreign trainers in local institutions, has been too piece-meal and narrowly focused tohave much impact on the corporate culture of the local utilities or their cost structure. Onenotable exception has been Ghana's VRA, which has used a "twinning" arrangement with thesame foreign public utility for specialist advice and management services, including termassignment in Ghana, and training for its own staff in the foreign utility's country ever since itsestablishment, more than 25 years ago.
52
5.24 But among imported services, management contracts have the greatest potential to makea lasting impact on a utility. They are usually seen as complex, and believed to have littlechance of success since they are bound to be resented by the local staff. Experience in two ofthe countries under review does not support such apprehensions. In Burundi, for instance, in aperiod ofjust three years a foreign management group successfully realized considerableimprovement, including the transfer of certain functions to the private sector, accompanied bysubstantial staff reduction (Box 5.3). C6te d'lvoire provides an example of a longer-termmanagement contract with equally impressive results (Box 5.4). The partners in CIE, theprivate company that runs the power sector of C6te d'Ivoire, have drawn on their intimateknowledge of the local culture to improve overall performance. CIE's staff of 3500 employeesincludes less than I percent of expatriates. Its African manager considers that with appropriateadjustments this approach is replicable in other African countries.21
Box 5.3: Good Practice on Management Assistance and Staff Streamlining
The major problems of the Burundi electric power sector had been diagnosed asGovernment interference in day-to-day management, lack of management experience ofotherwise technically competent engineers, and reluctance of the Government to privatizeany activity. As a first step, a team of five expatriates was given executive responsibility andcharged with improving operating performance and training local managers. Each memberof the team had a specific expertise: human relations and capacity building (the teamleader); investment programming, economic justification, and tariff development; inventorymanagement; accounting and finance; and computerization of the management informationsystem. The exercise is considered a success. After only three years, the expatriate team hasleft the country and the management of the company is firmly in the hands of localprofessionals.
During this three-year period, personnel was reduced from roughly 1,500 to 1,000without any noticeable social problem. This was done by ensuring that the staff released wasplaced in the private firms which began to carry out work heretofore done in-house, such asthe production of wooden poles, the connection of new customers, etc. Another majoraccomplishment was the reduction of accounts receivable other than from the public sectorfrom about 9 months of sales equivalent to less than 3 months.
21. Zadi Kessy, 'Presentation d'une Methode de Management appliqu&e a la Compagnie Ivoirienne d'Electricite". Abidjan, April1994.
53
Box 5.4: Good Practice on Management Contract ("Affermage")
Until the 1980's, the electric power sector of C6te d'lvoire had been operated relativelyefficiently under a concession contract with EECI, a company majority-owned by theGovernment. However, during the decade that followed, the sector experienced a generalif gradual performance deterioration. A combination of low operating efficiency and weakfinancial management brought the sector into progressively deeper financial difficulties,eventually leading to its inability to service its debt. Typically, physical and administrativepower losses increased, collection performance deteriorated, and tariffs reached a levelwhich could no longer be justified except to make up for sector inefficiency. In the late1980's, DCGTx, a department of the Prime Minister's Office charged with the supervisionand control of the major investments of state-owned companies presided over thereorganization of the electric power sector. As a result, EECI retained the responsibilityfor sector planning and investments, including the ownership of sector assets on behalf ofthe Government, while CIE, a private company created for this purpose in October 1990,was awarded a management contract for operation and maintenance. The two mainshareholders of CIE are the French SAUR and EDF. It is important to note that SAUR isalso one of the major shareholders of SODECI, the successful private company which hasbeen responsible for the operation and maintenance of Ivory Coast's urban water supplysince 1959. Not surprisingly, CIE has substantially adopted the modus-operandi andcorporate culture of SODECI. In the event, CIE is very decentralized, has elaborateprocedures and descriptions of responsibilities, a good reward system, a strong budgetingand time management system, and offers training and other social support to its staff.Some 20 percent of its shares are quoted on the Abidjan stock market, and its employeescan buy shares at a preferential price. The results so far are impressive. Between 1989/90and 1993/94, power losses declined from 19.8 percent to 17.4 percent, while the averagenumber of hours of power outages per customer dropped from about 50 hours to 18 hours.Likewise, the number of employees per thousand customers dropped from 9.5 to 6.9 whilethe percentage of consumed energy actually billed and the collection of accountsreceivable increased from 94.8 percent to 97.3 percent and from less than 96 percent to98.3 percent, respectively. Some 2,000 employees out of a total of about 3,200 havereceived some form of training. To this end, CIE spends the equivalent of 4 percent of itstotal annual salaries.
Commitment Lending
5.25 The new policy leaves no room for complacency. Yet, it is not couched in "absolutes"; itonly requires substantive progress towards the objectives embodied in the first three "principles"of the new policy. A review of ten Bank-financed projects in SSA, five of which were approvedin 1992, four in 1993 and one in 1994 shows that a few are still designed in a rather traditionalway (Togo/Benin, Malawi). However, the vast majority of the loans include elements of the newBank policy.
* In Zimbabwe, an operational plan is aimed at changing the legal and regulatoryframework, to increase the autonomy and complete the commercialization of thepower companies.
54
* In Guinea, a new regulatory framework and the restructuring of the power company,ENELGUI will allow the transfer of operations management to a private group.
* In Rwanda similarly, sector operations will be contracted out following the creationof a mixed company in which the government will own no more than 20 percent ofthe share capital.
* In Sierra Leone, a performance-based management contract will lay the basis for anautonomous and commercially viable entity.
* In Angola, a new institutional and legal structure will be created, including theseparation of regulatory functions from the administration and management of theutilities, and the development of commercially viable and autonomous powercompanies.
* In Ghana, the commercial operations of ECG, the power distribution company, willbe carried out by a private firm under a performance contract. (Box 5.5)
* In Tanzania, the power company will offer for sale its in-house transmission anddistribution networks construction activities and study other privatizations options.
Box 5.5: Good Practice on Commercialization Objectives
In Ghana the ultimate objective is a financially viable, well-managed sector, whoseinstitutions enjoy a high degree of autonomy and are capable of delivering a reliable andeconomic supply of electricity. Specific measures identified to facilitate the achievement ofthese commercialization objectives, some of which were initiated under the NationalElectrification Project, include:
* "the development ... of a suitable power sector regulatory framework with the necessaryregulatory authority over tariffs and other related matters;
* creating ... a task force ... to evaluate options for improving the efficiency and efficacy ofthe ownership and corporate structure arrangements now in place in the utility sub-sector;
* maintaining the tariff making process on a sound technical basis using a formula-basedapproach;
* improving the internal efficiency of commercial operations in ECG by engaging acontractor on the basis of performance-based management contract to revamp billing,collection and customer relations (underway); and
* engaging a consultant to recommend a scheme of service for both utilities, includingmatters such as staff deployment, remuneration levels, incentives, sanctions, and anyother such matters likely to affect the motivation and efficiency of the utilities'personnel."
Source: World Bank Green Cover SAR, Ghana Thermal Plant Project, January 1994.
55
5.26 Up-Front and Disbursement Conditionality. Convincing SSA countries to change theirmodus operandi after condoning it, for decades in some instances, is difficult, and a good designdoes not guarantee commitment on the Borrower's part. After having accepted many breaches incovenants it seems that the Bank is now more cautious. In this respect, the following examplesof up-front conditions such as for negotiations, Board presentation and effectiveness is worthnoting. So are conditions of disbursement when the project components are appropriately phased.
* Conditions of negotiations were used to obtain a satisfactory proposal for technicalassistance with a mature utility (Sierra Leone), and for government to settle arrearsof certain public sector entities (Tanzania).
* One condition of Board presentation was used to ensure the settlement of publicsector arrears (Sierra Leone).
* Conditions of effectiveness were used for the signing of a performance-basedmanagement contract and the settlement of cross-debts (Ghana); the reduction ofstaff by 200 employees and the conclusion of the capital restructuring of the powercompany, the preparation of a regulatory and institutional framework, and theinvitation to bid for the management of the power company (Guinea); thecompletion of a new tariff structure, a 15 percent tariff increase, and the invitation tobid for the management of the gas company (Rwanda); the implementation of a 45percent staff reduction (Sierra Leone); the selection of external auditors (Angola); atariff increase, and the settlement of the debt of government agencies (Tanzania).
* Conditions of disbursements were used for the enactment of a regulatory frameworkand for holding sector assets, and the conclusion of arrangements for the transfer ofthe management of the power company to a private operator (Guinea); thepublication of the law establishing the regulatory and institutional framework forpublic enterprises, and the hiring of a private operator to run the gas company underlease contract (Rwanda); and hiring a private firm under performance-basedmanagement contract (Sierra Leone).
5.27 Supervision reports show "normal" implementation with a good deal of compliance, aswell as a few examples of delays in taking action. Angola offers an interesting lesson. The powersector rehabilitation project approved in May 1992 had to be modified in mid-1994 as a result ofthe war. Along with changes in the physical components, funds were reassigned to reinforce theinstitutional component by including a study to define a corporate strategy and for restructuringENE, the national power company. Even if already committed to change, the case of Angolashows the possibility of enhancing the move towards greater transparency and commercializationwithout having to wait for the "next" operation.
Private Investment
5.28 As with other infrastructure sectors, ownership of the power sector assets by privateentrepreneurs (assuming their interest could be elicited) may be a touchy political and economicissue in SSA, although privatization of some segments of the operations appear possible.
5.29 Private operators can be expected to invest in almost any country if they succeed incovering their perceived risks. Generally, the higher the risk, the higher the return and the shorter
56
the pay-back period they will insist on. However, the return on investment is not the onlyconsideration. Guarantees also play an important role where the perceived risk is so high that theexpected return would exceed a certain level. Because of guarantees, deals are seldom if everconsummated with returns in excess of say 20 percent to 24 percent. Sometimes, an intimateknowledge of the market and long exposure to a country's culture suffice to sufficiently reducethe perceived risk. Such was the case of the recent creation of CIPREL (The Power ProducingCompany of C6te d'lvoire), a new private venture formed in CMte d'Ivoire by the two majorshareholders of CIE to construct and operate a combined-cycle gas turbine plant.
5.30 Implementation of the first three principles enunciated in the Bank's policy would go along way towards reducing the perceived risk to potential investors. Even in the countries thathave applied these principles, the risks perceived by would-be investors soon afterimplementation would still call for a premium and guarantees which governments may not wishor may not be in a position to entertain.
57
6. Conclusions and Recommendations
Bank policy for power lending makes good sense for SSA. Power lending shouldform part of acarefully balanced country assistance strategy including other energy resources such as biomassand be supportive of economic management objectives. Large loans should be avoided whereeither performance is below acceptable benchmarks in key technical andfinancial areas or noacceptable platform of sector reforms is underway. The Bank should work at increasing theownership in SSA countries of its policy objectives and principles, TA programs and the modalityof TA delivery. It shouldforge a partnership with other donors to obtain a consensus on criteriagoverning their involvement in the SSA countries and to improve the deployment of resources.
The Bank should be selective when promoting restructuring and privatization. Enablingconditions for competition in generation are absent in SSA, but an adequate tarifffor small andmedium bulk purchases should be put in place everywhere. Decentralized community-basedsupply which has worked well in other sectors is worth exploringfor low voltage service.Commercialization of distribution is key. Unbundling it from transmission and splitting it intofranchise areas would open the fieldfor private operators to compete for management contracts.Privatization is controversial but probably necessaryfor investmentfunding In Cote d'Ivoire,
the IPP for new thermal capacity and the amendment of the management contract into aconcession is a good illustration of issues and feasible solutions.
6.1 Much remains to be done to ensure that utilities operate in a business-like fashion. Butsome countries are showing that it can be done in SSA. This concluding chapter makessuggestions for the task ahead. These suggestions fall into two categories: first, a commonmenu for all countries and, second, suggestions about sector restructuring and private sectorfinancing, the "new frontier" of institutional development. The chapter then outlines what couldbe the role of the Bank in helping implement this agenda. It closes with a few recommendations.
The Road Ahead: Immediate Agenda
6.2 Whether or not utilities remain vertically integrated, attention continues to be needed onfour fronts: sector policies, regulatory principles, investment program, and human resources.
6.3 Sector policies. Some tenets must be enforced with vigor, often with the help of specificlegislation:
(i) Governments should concentrate on their roles as regulators and policy makers andrefrain from any other form of control on the sector.
(ii) Power utilities should operate on the basis of clear objectives enshrining soundtechnical and commercial principles, including the possibility to spin off in-houseservices which can be contracted out.
58
6.4 Regulatory principles should strive to ensure sector efficiency and financial viabilitythrough the following:
(i) Promoting efficiency and equity through:
* a considerable reduction, if not complete abandon of subsidies andcross subsidies
* a shift from cost-plus based regulation to incentive-based contracts;
* ensuring that contracts are awarded through a transparent andcompetitive process.
(ii) Enforcing certain rules of financial discipline on account receivable by private andpublic sector, self financing (30 to 60 %), returns on assets (8 to 12%); transparentcompensation by Government for any non commercial obligation that it imposeson the utility such as subsidized service to rural areas.
6.5 Investment program. Power sector investment must be limited to the financial leveragecapability of the sector. Its composition should be prioritized, initially to rehabilitate generationand distribution, and later to expand distribution and generation. The internal economic rate ofreturn of the investment program and of Bank financed projects should be greater than 10percent when benefits are calculated at the prevailing tariffs.
6.6 Human resources. The availability of abundant and skilled human resources and thecapability of institutions to put it to work, are key to social and economic development. For thepower sector, the implementation of a comprehensive sector manpower development program atall level is necessary for both the utilities and the government. The integration of expatriates,and twinning arrangements with well functioning utilities can initially help both to fill skill gapsand transfer skills to local employees. The task is no less challenging and largely in unchartedwaters when it comes to training government in fulfilling a more sharply defined, and morelimited role as sector regulator. Finally, incentive systems must be created to help power utilitiesand the Government institutions make the most of new skills acquired by managers and staff.
The New Frontier: Restructuring and Assets Privatization
6.7 With few exceptions, the power utilities of SSA countries are government-ownedintegrated monopolies responsible for transmission and distribution of all electricity that theyproduce. Varying degrees of competition exist in the running of even the most integrated powerutilities, as for example to construct facilities, to contract out certain functions, etc. One of thetenets of the new policy is that exposure to markets and competition should be increased in orderto enhance performance. In particular, recognition that not all of the activities involved in theprovision of infrastructure services are a natural monopoly is central to achieving greaterefficiency.
6.8 Competition in generation. At this stage, the small size of the power systems and theweakness of the regulatory frameworks of SSA countries make it difficult to recommend theirbreakup (unbundling) as was done in some countries like the United Kingdom and Chile.
59
Unbundling of generation from transmission and wholesale competition for generation is acomplex affair. One consequence is the proliferation of contracts both in the short and long termnot unlike those found on spot and future commodity markets; another one is the fragmentationof the generation system into several pieces, for competition to really emerge, and a trendtowards smaller plants and less capital intensive generation.
6.9 This model also assumes that economies of scale in generation can be obtained in spiteof sector fragmentation or are less important than the efficiency brought by competition. It alsoimplies a streamlined but sophisticated regulatory framework, commercial skills in abundancefor contracts management and above all a strict legal framework to enforce them. Absent these,as they are in most SSA countries, producers will minimize transaction costs and the risks of nothaving contracts enforced with cartelization and vertical integration (whether tacit orinstitutionalized); in essence transaction costs can make a monopoly natural even wheneconomies of scale do not. Besides, such economies exist in SSA countries where large scalehydro or coal-based power is least cost (para. 3.6); from that point of view, competition on abroad scale and system fragmentation are desirable only in systems where small diesel sets orgas turbines are economic over the long term e.g. in landlocked or gas producing countries.
6.10 Independent producers might be able to compete with utility-based generation in specificcircumstances, so there is a need to review the barriers to their entry in the market and the rulesfor pricing their output. To start with, a marginal cost-based tariff for all small and medium bulkpurchases should be put in place. Thus, most SSA countries would continue to have a dominantbut not unique generating company until conditions for a fully competitive market are met.
6.11 Decentralization of distribution. The weaknesses that prevail at national level in theformal governance system of SSA countries suggest another avenue, that of less sophisticateddecentralized community-based supply. Whether in education, health, water supply andsanitation and rehabilitation or supply of new housing, this mode of service delivery has elicitedresponsible behavior and good cost recovery from beneficiaries, and responsive supply byproducers. However, these promising results were generally obtained for technologies with noeconomies of scale and little demand for specialized technical and managerial skills.Decentralized community-based low voltage power supply is worth exploring, but it shouldexclude most traditional generation technologies except minihydro. Decentralization has onedrawback in the eyes of many: it often translates into markedly different power prices from onearea to the next, a significant departure from the prevailing practice of a national tariff for smallconsumers.
6.12 Unbundling of distribution. Restructuring distribution has attracted less attention thancompetition in the generation market. This is because the business orientation of distribution istaken for granted in industrialized countries. Distribution should be the prime focus of attention:it is where service delivery and cost recovery happens. In SSA countries, whateverimprovements are made in efficiency at plant and grid level would be jeopardized if distributionremains grossly overstaffed, unreliable and does not collect bills. Management contracts canmake a big difference, but a major issue can be the scarcity of bidders or the dominance of a fewones as has happened in water supply in some countries. Unbundling distribution fromtransmission and generation opens the field for private operators which would not have thetechnical and financial wherewithal to invest large sums. It has only one drawback: distributorsmay have to set prices above efficient marginal cost-based tariffs in order to remain financially
60
afloat. Unbundling of distribution should be accompanied by its decentralization into severalservice areas designed to make the bidding for the franchise attractive.
6.13 Privatization. In LDCs, the share of privately owned power sector assets is increasing asa result of two movements: the divestiture of existing assets by government mostly in LatinAmerica and the emergence of Independent Power Producers (IPPs) mostly in Asia. In SSA,only the latter is happening, and it is happening only in Cote d'Ivoire.
6.14 An OED evaluation of completed privatizations22 shows that ownership of privatizationprograms by SSA countries was generally lacking even in the industrial and agricultural sectors.The issues of asset valuation, fair price, labor redundancy and statutes and need for governmentcontrol often lie at the heart of these feelings. In addition, assets privatization in the energy sectorraises specific problems such as the discomfort of having a private monopoly when regulatorycapacity is known to be weak, the transfer price of national resources such as hydropower andnatural gas which can generate a rent for its producers. All these issues can in principle besolved-through the setting of royalties and strong regulation-but their practical solution is likelyto be controversial and protracted. It is likely to be least controversial for new thermal power and,if labor problems can be overcome, for distribution. The problem has often been eased in Frenchspeaking countries by the signing of build, operate and transfer concessions for a limited time afterwhich asset ownership is transferred to government.
6.15 Private ownership of assets is probably necessaryfor investmentfunding. Though lesspressing a problem than in booming economies, investment funding will soon become so in SSAcountries where growth for effective demand is picking up and cannot be met by efficiency gains.Lessons from the past are clear: first, governments are not prone or financially capable to be goodshareholders; second, management contracts, which by definition do not require operators tofinance investments, are no guarantee of adequate cost recovery. In many ways, Cote d'Ivoire isshowing both the contours of the issue and possible solutions for SSA countries: arrears are lowenough, efficiency and power prices high enough for CIE to meet its mandate, but this is notsufficient to finance system expansion. Even EECI is unable to fund new connections and minoroverhauls of the distribution system. A recent amendment to CIE's management contract extendsCIE's responsibilities to investments thus making it evolve towards a concession contract. COted'lvoire is also mobilizing private equity with an IPP contract for gas-based generation.
6.16 Reform agenda items are interdependent. Privatization is often not easy, and in the powersector of SSA perhaps not possible, until physical rehabilitation and/or organizational restructuringhas taken place. Privatization, on the other hand, reduces the need for the regulator to enforceminimum financial targets while organizational restructuring increases the regulatory burden as thenumber of entities and pricing arrangements multiply with unbundling and decentralization, clearlya serious constraint for SSA countries to consider in setting the pace of reforms. Although sharingsimilar cultural traditions, history evolution, and several common problems, each country shouldselect its own institutional options23 for the power sector and design its own program.
22. World Bank Assistance to Privatization in Developing Countries, Report No. 13273, August 19, 1994.23. Country parameters influencing the choice of options are: GDP level and growth rate; country size; population size,distribution, density and growth rate (number and size of urban and rural centers); population literacy. Power sector parametersinfluencing the choice of options are: Demand size and growth rate (MW; electricity consumption per capita and growth rate; shareof electricity consumption by industry sector; endowment in energy resources (hydro, gas, coal, petroleum).
61
6.17 One of the major determinants of the extent of change sought by governments will be thelevel of awareness among the stakeholders (public officials, political leaders, consumers andinterest groups) about the "correct" behavior in the more successful countries. Another will betheir perception that other countries at similar stages of development in the region or in otherparts of the world have benefited from applying the principles which the Bank is promoting.Without that awareness and perception, the status-quo may well prevail. Action programs axedat enlightening the stakeholders leadership of those countries would appear to be essentialingredients for a successful transition to a more business-like environment.
Role of the Bank
6.18 The Bank, other multilateral institutions and the bilaterals have had in the past animportant role in assisting the countries build up the sector and overcome energy crises. TheBank, the UNDP and bilateral lenders have also made important contributions in the EnergyAssessments and ESMAP initiatives. However, judged from sector performance in the early1990's, their institutional development impact has been mixed. Together, donors could exercisemuch more leadership and achieve more than they have in the past in helping the SSA countriesimplement the agenda presented in the above paragraphs.
6.19 The Bank comparative advantages are (a) the financial resources that it can provide; (b)its knowledge and experience in infrastructure-related issues; (c) its capacity to pool information;(d) its capacity to insert a power project in a broad policy framework; and (e) its influence inadjustment policies through its lending and technical assistance operations. The Bank hashowever strong competitors on (a) and (b) and is increasingly seen as a knowledge and policyadvice Bank. It should cultivate this relative strength in many SSA countries.
6.20 Bank policy for power lending makes eminent sense for SSA. The poverty of SSA an itsdependence on foreign assistance do not justify relaxing quality standards. Indeed powerutilities are peculiarly vulnerable to the dysfunctions of public management and poorlyperforming utilities in turn have a depressing impact on the countries fiscal performance, alreadya very weak spot of many SSA economies.
6.21 Power lending shouldform part of a carefully balanced country assistance strategy.The Bank has a strategic interest in the sector but it would be counterproductive for the Bankand the borrowers to pour money into unsustainable power sector expansions. The priority givento power lending can be questioned where borrower ownership of sector policy reform is notassured: power infrastructure does not trigger growth on its own and poverty alleviation as wellas important energy and natural resources management objectives are best served by thedevelopment and rational use of traditional energy resources such as charcoal, fuelwood andother biomass. Because it is capital intensive, power lending should be especially supportive ofeconomic management objectives (fiscal restraint, efficient use of financial and human capital bypublic and private sectors) and complement adjustment operations; it should not be viewed as avehicle for transferring resources or for targeted subsidy interventions.
6.22 Signs exist that the Bank is well on its way to realigning its sector strategy in SSA.Operations dealing with the management of forestry resources, fuelwood distribution andimproved stoves are picking up in scope and number. This reflects changes in the countries(decentralization of resources management) and experience accumulated by the Bank in manystudies and pilot projects. For example, projects in Mauritius, Mozambique and Niger will
62
introduce technologies on the transformation of biomass-bagasse, fuelwood-intobriquettes/charcoal, manufacture stoves and furnaces for household and industrial uses andcreate institutions to monitor preservation of biomass. Moreover, six on-going power projects inBurundi, Ethiopia, Malawi, Mali, Niger and Rwanda have one or more components for theproduction, marketing and distribution of charcoal and crop residues and efficient householdstoves.
6.23 The Bank is now applying, albeit still notfor all SSA recipients of recent loans, theprinciple of "commitment lending "enshrined in the new policy. It is encouraging to note theincreased use of up front conditionality for larger operations addressing system expansion and tosee these conditions associated with measures which would have appeared as too radical only afew years ago e.g. management contracts with private and mostly off-shore operators in half adozen cases.
6.24 The Bank's continuing emphasis on physical andfinancial rehabilitation is well placed.It allows for very relevant interventions which are modest in size; such operations can help buildup the capacity and test the commitment of borrowers in tackling increasingly challengingobjectives on the road to power sector reform.
6.25 Few of the SSA countries have reached a stage where the Bank can and should inducelarge scale private investment in the power sector by providing guarantees. The cost of theseguarantees would end up as unacceptably high compared to the benefits as long as the policyrisks are still very substantial or perceived as such. However, a few such operations may still bedesirable in SSA countries which stand most to gain from IPPs involvement as in Cote d'lvoireor Tanzania to develop gas-fired generation. In most other cases, the Bank group can assist inmany ways as it did in other parts of the world: advice and training for project identification,evaluation and contract bidding, helping governments to obtain alternatives to sovereign or Bankguarantees, equity or syndicated debt financing by IFC.
6.26 New style interventions necessitate major inflections in the Bank role and modusoperandi, such as:
* the nurturing and gauging of borrower's ownership of the project at every step and inall its dimensions: locus of initiative, intellectual congruence, political commitmentand stakeholders' persuasion.
* a better agreement and coordination with other donors on each country strategy.
* for smaller and TA focused operations, a better deployment of Bank staff,consultants and division of labor among donors with experience and presence in thefield.
6.27 The Bank can assist the countries in building up awareness of the needfor sectorreforms by: having dialogues with the Government, management and staff of sector entities andinterest groups (NGOs, industrial consumers, community consumers); sponsoring regional ornational seminars and workshops on successful cases appropriate to the particular conditions ofthe countries. Such help was provided to Latin America (Cocoyoc Conference in Mexico,Seminars in Colombia and Bolivia in the early 90's), to Eastern Europe and former Soviet Unioncountries (1993 and 1994 Workshops in Vienna, Austria), and India (1993 Conference).
63
6.28 The Bank has accumulated valuable knowledge through studies of the energy sector andof the power subsector in the SSA countries which provided a sound basis for policy choices anddonors involvement. The execution of this evaluation has shown that infornation on the powersector is often scant, inconsistent and not adapted to today's needs. In the changing political andeconomic environment of the region, improved data gathering and analysis is needed to monitorand evaluate sector performance and to provide a basis for private sector involvement.
6.29 Finally, the Bank also has relevant and fresh experience in forging strategic partnershipswith other donors with respect to infrastructure development-e.g. the Road MaintenanceInitiative, a five year program carried out as part of the SSA Transport Program initiated in 1987and which has already yielded rich lessons.
Recommendations
6.30 The new policy orientations for power lending by the Bank makes sense for SSA-itspoverty and dependence on foreign assistance do not justify relaxing quality standards. Indeed,power utilities are peculiarly vulnerable to bad management, and their poor performance is acontributing factor for the weak fiscal situations in many SSA economies.
6.31 The Bank and other co-financiers could exercise much more leadership in promoting thesector reforms that form the thrust of the new agenda. Until such reforms take hold, the Bankmust exercise caution in providing guarantees to induce large investments in power sector.
6.32 Based on the foregoing findings, this review has six main recommendations.
Recommendation 1. The Country Assistance Strategies for SSA countries shouldexamine the justification for power lending and establish its coherence with Bankassistance for broad economic adjustment and for the development and rational use ofother (especially renewable) energy resources (paras. 4.6-12 and 4.25-27).
Recommendation 2. Except for small operations aiming at institution building orrehabilitating facilities (para. 6.24), power lending in SSA should be avoided incountries where sector performance is below acceptable benchmarks in key technical andfinancial areas24 and where little of the following reform platform is underimplementation at the time of appraisal (paras. 5.9-22 and 5.26):
* The establishment of a transparent and arms-length regulatory frameworkwith legal guarantees that utilities can operate with autonomy-for example,through management or concession contracts; and
* The enforcement of regulatory principles to ensure financial discipline,adequate tariffs, and incentive-based, competitive contracting of services.
Recommendation 3. When promoting power sector restructuring and privatization inSSA, the Bank should explore setting up purchase tariffs, decentralizing distribution andunbundling it from generation and transmission, using concession contracts for private
24. Areas to monitor are technical losses, accounts receivable and rate of return on investments.
64
operators and providing guarantees for independent power producers (pams.6.8-15).
Recommendation 4. The Bank should nurture SSA borrower ownership of its newsector policy principles, of institutional development programs to support the reforms;and of the delivery of technical assistance by locals and expatriates. Effectivedissemination of good practice and build-up of stakeholders' support should be anintegral part of project preparation (pams. 4.13-24 and 6.27).
Recommendation S. The Bank should forge strategic alliances with other lenders anddonors to obtain a consensus on the policy objectives and criteria for their involvementin SSA countries. It should also establish partnerships in the deployment of humanresources and share responsibilities in performing those tasks that would benefit fromthe diverse field assets and competencies of donors (paras. 4.4450).
Recommendation 6. In collaboration with other interested lenders and donors, the Bankshould help coordinate and institutionalize a systematic effort to gather and analyze dataon the power sector in line with the principles of the new power sector policy (pams.3.16-19, 5.18, 6.19 and 6.28).
Projects in The Study Cohort
Cosin~try 1./C No. Prqj. Name FY Major Sector UiC.. mnl. Tot. D04b. Tat. TotU Ovra u Sut ' hst. Dev.Counuy /C MO.png. Nw FY mom secto uc 4.nil. $ WA CceL*. UndlabI. $IRv Re" RtinthMo. WA.
BOTSWANA L22090 POWER I 1983 Thermal 32.50 18.23 14.27 0.00 s Iik par
BURUNDI C15930 TRANS & DISTR.SYSTM. 1985 Dint/Trarsm 12.30 11.59 3.84 0.00 s Ik par
EASTERN AFRI C14190 RUZIZI/ZAI/RWA/BUI 1984 Hydro 15.00 16.56 0.10 0.00 u unc par
EASTERN AFRI C14200 RUZIZI/ZAI/RWAMBUI 1984 Hydro 15.00 16.55 0.11 0.00 u unc per
EASTERN AFRI C14210 RUZIZIlZAI/RWA/BUI 1984 Hydro 15.00 16.54 0.11 0.00 u urnc pr
GHANA C16280 POWER 1986 DistrlTransm 28.00 34.87 0.00 0.00 a Iik sub
GHANA C17590 NORTH GRID EXTENSION 1987 Hydro 6.30 8.88 0.00 0.00 a unc sub
GUINEA CS0220 POWER ENGINEERING 1980 Pow rSector 1.13 0.00 1.13 1.13 u unl par
GUINEA C10850 POWER I 1981 Power Sector 28.50 23.68 0.00 0.00 u unl neg
GUINEA Cl 5950 POWER ENGINEERING 11 1985 Power 8.00 10.95 0.05 0.00 u unc neog
COTE DIVOIRE L18960 POWER I 1981 Power Sector 33.00 33.00 0.00 0.00 s uric sub
COTE DIVOIRE L31500 ENERGY SECTOR LOAN 1990 Power 100.00 100.00 0.00 0.00 s lik par
KENYA LS0120 ENGR. LOAN/OLKARIA G 1979 Power Sector 9.00 8.40 9.00 0.00 s lik sub
KENYA L17990 POWER-III 1980 Power Sector 40.00 38.70 1.30 0.00 a lik sub °
KENYA U22370 POWER IV OLKARIA 3RD 1983 Power Sector 12.00 7.46 4.54 0.00 s Iik par
KENYA C14860 ENERGY EGR GEO 1984 Oil & Gas 24.50 30.04 0.00 0.00 s lik par
KENYA L23590 POWER V (KIAMBEREI 1984 Power Sector 95.00 79.00 16.00 0.00 s urnc neg
UBERIA L16000 POWER IV 1978 Power Sector 10.00 10.00 0.00 0.00 s unc par
MADAGASCAR C08170 POWER IIANDEKALEKA) 1978 Hydro 33.00 33.00 0.00 0.00 u lik par
MADAGASCAR C08171 POWER l(ANDEKALEKA) 1978 Hydro 10.00 7.57 2.43 0.00 u lik par
MAU C12820 POWER/WATER 1983 Power Sector 24.00 23.36 0.01 0.00 u unc neg
MAURITIUS L15480 POWER II 1978 Power Sector 15.00 11.10 3.90 0.00 5 lik par
NIGER C15110 POWER 1985 Power 7.50 8.72 0.18 0.00 s lik sub
NIGERIA L17660 POWER V 1980 Distr/Transm 100.00 100.00 0.00 0.00 u unl neg
NIGERIA L20850 POWER VI 1982 Distr/Transm 100.00 100.00 0.00 0.00 u unl nag
RWANDA C14950 TRANSM.& DISTRI.SYST 1984 Distr/Transm 9.00 10.05 0.00 0.00 u unc par
SENEGAL CS0260 POWER T. A. & ENGR. 1980 Power Sector 3.30 3.30 0.00 0.00 s lik par
SENEGAL C17100 ENERGY SECTOR 1986 Power Sector 20.00 22.16 1.50 0.00 u unc nag
SIERRA LEONE C07340 POWER III 1978 Thermal 8.20 8.19 0.01 0.00 5 unc neg x
Tot. Diab rot. Total Overall Sust. Inst. Dev.Country LUC No. Proj. Name FY Major Sector L/C S. mil. S; m Cancel. Undiab. $, Rating Rating Rating
*ml Mol. mil.
SIERRA LEONE C12650 POWER ENG. CREDIT 1982 Power 5.00 5.13 0.00 0.00 u unc neg C
SWAZILAND L20090 POWER III 19B1 Thermal 10.00 10.00 0.00 0.00 s lik par ;.
SWAZILAND L20091 POWER III 1981 Thermal 5.60 5.60 0.00 0.00 s lik par
TANZANIA C14050 POWER IV 1984 Power Sector 35.00 33.32 2.88 0.00 s lik par
TANZANIA C16870 POWER REHAB./ENERGY 1986 Power Sector 40.00 44.27 1.40 0.07 s unc neg
UGANDA C15600 POWER 11 1985 Power Sector 28.80 40.17 0.00 0.00 s unc per
WESTERN AFRI C 11890 NANGBETO ENG. CREDIT 1982 Power 1.80 1.70 0.10 0.00 s lik sub
WESTERN AFRI C11900 NANGBETO ENG. CREDIT 1982 Power 1.98 1.85 0.06 0.00 s lik sub
WESTERN AFRI C15070 NANGBETO 11 1984 Hydro 15.00 17.73 0.00 0.00 s lik sub
WESTERN AFRI Cl 5080 NANGBETO 11 1984 Hydro 15.00 17.73 0.00 0.00 s lik sub
ZAIRE C12240 REHAB.SHABA PWR.SYST 1982 Power Sector 19.00 17.65 0.00 0.00 u unc par
ZAIRE C17120 POWER 11 1986 Power Sector 37.00 50.33 0.00 0.00 u unc par
ZAIRE C22930 SNEL TA 1992 9.89 7.51 2.82 0.00 u unc par
ZIMBABWE L22120 POWER I 1983 Power Sector 105.00 105.00 0.00 0.00 s lik sub
ZIMBABWE L29000 POWER 11 1988 Power Sector 44.00 42.52 0.00 1,48 s* * Ch~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
On-Going Projects
Total Disbur. Total Cancel. TotalCountry UC No. Proj. Name FY Major Sector LIC S mil. T mol D, mlC. Unsb
S. mil.
AFR. REGION C23660 TOGO/BENIN ENGINEERING 1992 2.75 0.63 0.00 2.38
AFR. REGION C23670 TOGO/BENIN ENGINEERING 1992 2.75 0.63 0.00 2.38
ANGOLA C23850 POWER SYSTEMS REHAB 1992 Energy 33.50 1.31 0.00 34.08
BENIN C22840 POWER REHAB 1991 Distr/Transm 15.00 1.59 0.00 14.37
BURUNDI C22300 ENERGY SECTOR REHABILITATION 1991 Energy 22.80 6.10 0.00 17.22
C.A.R. C19780 ENERGY 1989 Power Sector 18.00 15.93 0.00 2.48
ETHIOPIA C17040 ENERGY 1986 Power Sector 62.00 49.55 0.00 24.28
GHANA C20610 POWER SECTOR(ECG FIFTH POWER) 1990 Distr/Transm 40.00 24.30 0.00 18.04
GHANA C21090 VRA/SIXTH POWER 1990 Hydro 20.00 4.32 0.00 17.14
GHANA C24670 NAT'L ELECTRIFICATION 1993 DistrfTransm 80.00 0.17 0.00 77.83
GUINEA C24160 POWER II 1993 Hydro 50.00 10.35 0.00 41.07
KENYA C19730 GEOTHERMAL DEVELOPMENT 1989 Power Sector 40.70 35.03 0.00 8.76
MADAGASCAR C17870 ENERGY I 1987 Power Sector 25.00 21.67 0.00 5.62
MALAWI C19900 ENERGY I 1989 Hydro 46.70 32.68 0.00 15.60
MALAWI C23860 POWER V 1992 Power Sector 55.00 1.87 0.00 53.49
MALI C19980 POWER II 1989 Power Sector 33.00 7.95 0.00 25.97
NIGER C18800 ENERGY 1988 Distr/Transm 31.50 11.89 0.00 22.44
NIGERIA L31160 POWER MAINTENANCE/REHAB 1990 Distr/Transm 70.00 37.81 0.00 32.19
RWANDA C24560 ENERGY SECTOR 1993 Power Sector 26.00 0.85 0.00 25.28
SIERRA LEONE C23560 POWER REHAB 1992 Distr/Transm 21.00 1.19 0.00 20.43
TANZANIA C24890 POWER VI 1993 Power Sector 200.00 1.29 0.00 202.48
TANZANIA C23300 ENGINEERING CREDIT 1992 10.00 7.02 0.00 3.69
TOGO C21710 POWER REHABILITATION 1991 Power Sector 15.00 5.30 0.00 10.66
UGANDA C22680 POWER III 1991 Power Sector 125.00 21.17 0.00 101.43
ZIMBABWE L36960 POWER III 1994 Power Sector 90.00 0.00 0.00 90.00
W.
