E N E R G Y & M I N I N G S E C T O R B O A R D D I S C U S S I O N P A P E R S E R I E S P A P E R N O . 2 J U N E 2 0 0 2 Global Electric Power Reform, Privatization and Liberalization of the Electric Power Industry in Developing Countries R. W. Bacon and J. Besant-Jones THE WORLD BANK GROUP The Energy and Mining Sector Board AUTHORS ACKNOWLEDGMENTS We would like to thank Renier Lock, Bernie Tenenbaum, and a referee for their helpful comments. We would also like to thank Cameron McKenna for financial support that made possible the study shown in Table 5. DISCLAIMER The findings, interpretations, and conclusions expressed in this study are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. CONTACT INFORMATION To order additional copies please call the Energy Help Desk. 202-473-0652 energyhelpdesk@worldbank.org AUTHORS R.W. Bacon rbacon@worldbank.org; J. Besant-Jones jbesantjones@worldbank.org This Paper is available online www.worldbank.org/energy/ The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy- ing, recording, or inclusion in any information storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copyright Clearance Center, Inc, 222 Rosewood Drive, Danvers, MA 01923, USA, fax 978-750-4470. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street N.W., Washington DC, 20433, fax 202-522-2422, e-mail: pubrights@worldbank.org. E N E R G Y & M I N I N G S E C T O R B O A R D D I S C U S S I O N P A P E R S E R I E S P A P E R N O . 2 J U N E 2 0 0 2 Global Electric Power Reform, Privatization and Liberalization of the Electric Power Industry in Developing Countries R. W. Bacon and J. Besant-Jones The World Bank, Washington, DC THE WORLD BANK GROUP The Energy and Mining Sector Board Copyright © 2001 The International Bank for Reconstruction and Development/The World Bank. All rights reserved ABSTRACT This paper reviews the progress of the movement to privatize and liberalize the power sector in developing countries. It reviews the forces driving the movement and then describes the steps that should be taken to achieve success. Data on actual steps taken and preliminary information on the impact of reform are presented. Finally, lessons from this past experience are highlighted. INTRODUCTION In the past decade, the pace of reform and change in the to improve performance, eventually resulted in excessive costs, electricity sector has rapidly increased, and the nature of the low service quality, poor investment decisions, and lack of reforms adopted have become steadily more sophisticated. innovation in supplying customers. 1 Many countries--from the very large, such as China, to the very small, such as Bolivia--have enthusiastically adapted Rapid changes in technology have occurred in both the earlier reform models to their own needs and circumstances. generation of electricity and in the computing systems used Both developed and developing countries have embarked on to meter and dispatch power. These changes have made new a program of liberalizing and reforming their power sectors. industrial structures possible, which state enterprises have been too slow to adapt. The private sector offers many new The principal driving forces behind this reform movement, solutions to providing power at lower cost, especially to described by a number of authors (1,2,3,4), include the consumers with low levels of demand, through innovations in following: (a) the poor performance of the state-run electricity customer service (service standards, billings, and collections) sector in terms of high costs, inadequate expansion of access and cost recovery mechanisms. These solutions are important to electricity service for the population, and/or unreliable requirements for the sector. supply; (b) the inability of the state sector to finance needed expenditures on new investment and/or maintenance; (c) The principal gains hypothesized as arising from sector the need to remove subsidies to the sector in order to release reform come from three separate sources of improvement in resources for other pressing public expenditure needs; and (d) economic performance. First, in terms of overall allocation of the desire to raise immediate revenue for the government resources, making consumers pay at the margin what it costs through the sale of assets from the sector. to produce and supply them is expected to achieve a better economy-wide use of resources. Issues of income distribution In many countries all these factors have been present at the and support for the poor are increasingly regarded as being same time, with the notable exceptions of countries in Eastern supportable by targeted subsidies to needy groups, rather Europe and the former Soviet Union, where policies of than by across-the-board subsidies, which have the effect of encouraging heavy industrialization had left the power sector generally distorting patterns of the consumption of energy. The with short-term excess capacity, so that new capacity was a extraordinary levels of subsidies seen in some countries (5,6) lower priority than in most other countries. have been calculated to produce major welfare losses in terms of overall economic welfare. Second, the profit motive gives a Although some traditional state-owned and -run enterprises stronger incentive for efficient use of inputs, in terms of lower have performed well, and indeed were often formed by cost combinations of inputs and actual reductions in inputs nationalizing private-sector companies that were either too required to produce a given output, than any incentives small to exploit economies of scale or too large to prevent offered by an enterprise controlled and managed by a monopolistic abuse, there was an increasing awareness during bureaucracy (7). Third, competition, where it is possible, is the 1980s that a lengthy period of state ownership, without the form of private participation most likely to reduce the costs the forces of competition or the incentives of the profit motive of production and to pass benefits on to consumers. If the sector can be made to cover its costs and be profitable, then finances through the asset sale (as was partially the case in there will be an incentive for firms to invest, and they will the United Kingdom). Because there are always interest groups also have an incentive to seek out new markets that can be likely to lose out as a result of sector reform--such as those profitable. New entrants, also attracted by profit opportunities, employees of the former state-owned enterprise who will be can seek out specialty market niches that may not appeal to made redundant, and those bureaucrats and politicians who mainstream firms. will lose a sphere of patronage--any such groups with political power must see sufficient benefits to outweigh the Although many countries have expressed some dissatisfaction costs in order to support the reform. with the operation of their state-owned power sector, there has been a wide range of responses to the problems perceived. Even when it has become evident to the ruling party that an Some countries have felt it impossible or undesirable to embark institutional change would be beneficial, the government must on any reform strategy that entails opening electricity production have the confidence that the reforms, several of which are or sales to private participants, whereas other countries, likely to require legal and, in some cases, constitutional although willing to engage private participation, have chosen changes, are politically feasible. Here, the strength of the very different strategies for doing so. The variety of responses majority, the nearness of the next election, and the mandate that have already emerged globally is one of the most striking of the previous election all impact on the willingness and 2 features of the power sector in the past decade. ability of the government to institute the required changes. A crucial window of opportunity may be created by a change There has been considerable interest in the wider issues of why of government because the incoming group may have the some countries chose to undergo economy-wide reforms in mandate, strength, and time to carry out the program. In terms of reducing the role of the state, introducing the private many countries, although the problem and possible solutions sector into markets once exclusively reserved for previously state- became evident early in the 1990s, action was not possible owned enterprises, and generally liberalizing control (7,8,9,10). for several years, because of the political priorities facing the Both a wide-ranging analysis (7) and discussion (11) of these incumbent governments around that time. issues for developing countries suggest that two essential condi- tions must be met before reform is attempted. (a) It should be Intertwined with these country-related conditions are the actions generally perceived in the country that reform is desirable and (b) of the International Financial Institutions, which have been carrying out the reform agenda should be politically feasible. advocating and encouraging both macroeconomic and sector reform. Lending policies often have had a "carrot and stick" The desirability of reform focuses on the consequences of the structure, in that lending for institutional reform, which is often unsatisfactory performance of the economy or sector for those bundled with lending for investments to upgrade supply facilities who have political influence. For example, the failure to that are needed to support the reformed power market, will provide rural electrification would be seen as undesirable by attach conditions related to achievement of targets for release the large number of rural households in many developing of tranches of the loan (12,13). The extent that failure to countries, but usually this group is politically weak, a condition achieve the targets would jeopardize both disbursement of that in itself is not likely to persuade the government to tranches under the present loan and subsequent approval of change policies. At the energy-sector level the picture is more other possible future loans will also be a factor in determining complex--poor macroeconomic performance may indeed the country's commitment to the reform process. persuade the government to undertake sector reform (especially if the government were running an unfinanceable Sector problems in energy are most likely to be felt in terms of deficit and could no longer subsidize loss-making public nondelivery of the product. Power blackouts and brownouts enterprises). However, for reform to be embraced, it is likely are the most dramatic instance of this, with their very high that the sector itself would need to be underperforming in costs of alternative supply for those who have come to count some crucial way--both in terms of the delivery of energy to on the public supply of electricity. Quality of service, which important groups of users and in terms of its financial claims takes many forms, also can deteriorate and impact users on the central budget. A sector that was meeting all demands adversely. The failure of supply may be partly associated with and was not a drain on central government expenditure is not very low operating efficiency caused by lack of maintenance, likely to be seen as a high priority for reform, unless the theft, etc., and partly associated with lack of investment motivation is solely to solve a short-term problem of public caused by financial restrictions. The inability of a state enterprise (and eventually government) to finance new and needed ELEMENTS OF A SECTOR REFORM investment is often compounded by poor public-sector price or PROGRAM tariff setting, which does not allow the state owned enterprise to recoup all its costs, as well as by inefficiency in collecting all the In many developing countries, and in particular those in Asia, revenue due it. Hence, a strong hypotheses is that reform is the Middle East, and Africa, reform of the power sector starts more likely where there are obvious problems of shortage of from a market structure that is dominated by a state-owned supply, such as blackouts, and less likely where there is excess national power utility with a legally endowed monopoly and capacity, making financing investment less important. a vertically integrated supply chain encompassing power generation, transmission, distribution, and customer services. The political feasibility of sector reform is likely to be closely The rationale for this structure is minimization of the costs of related to the political feasibility of macroeconomic reform, coordination between these functions and of financing the but one possible difference is that sector reform is most likely development of power systems. The pre-reform structure in to eventually involve the privatization of existing assets, as well other countries, notably in South America, places distribution as private-sector involvement in new investment. In countries and customer services with local companies, separate from that have a relatively small, internal, formal financial structure national companies that provide power generation and (compared with the size of the sector) and possibly no stock transmission. market, privatization inevitably means foreign ownership in 3 part or in total. This raises different issues from those normally Power reforms are designed to introduce competition where associated with a macroeconomic stabilization. Ownership feasible, which is in the upstream production and downstream is seen as a long-term and irreversible change (although supply functions of the industry structure, and to use economic renationalization is not unknown--for example in the United regulation of the wholesale and retail power markets to Kingdom in the 1940s and in Latin America in the 1960s), promote competition and protect consumer interests. and the control of key domestic sectors by foreign companies Regulation of the power market is essential, as shown by the may be seen as qualitatively less desirable than the general, experience of New Zealand, which tried an approach without but temporary, austerity required by a stabilization package. the amount of regulation used elsewhere. Their approach Hence, governments willing and able to contemplate a strict was based on mandatory separation of generation, transmis- monetary and fiscal control, for instance, with its attendant sion, and distribution, using general competition laws to deal short-term recession may not be willing or able to contemplate with both the terms of interconnection and conduct generally privatization. In order for privatization to happen, the problem in unbundled power networks. Sector-specific regulation, must be more obviously linked to the underperformance of the especially of electricity prices, was rejected under this sector, and the government must be in a strong position, as approach as being self-defeating, and competition was relied expressed in terms of the support of those groups that are on to provide the required market discipline for participants. likely to determine its future. Experience, however, showed that competition was not sufficient to control pricing in the presence of the transmission The next section discusses the possible approaches to and distribution natural monopolies, and hence, the New power-sector reform: the various steps that can be taken and Zealand government subsequently imposed price controls on the sequence in which they should be taken. The following power suppliers (14). section gives a global perspective of progress with reforms to date in developing countries. This section looks at results from A full-scale power reform program generally consists of the steps already taken, as well as at patterns in the approach to following main elements. sector reform. Because the reform movement effectively started around 1990 (with the notable exception of in Chile, where it 1. Obliging electricity enterprises to operate according to started approximately 10 years earlier), there are now data commercial principles. This obligation extends to state-owned relating to the impact the electricity sector reform had on the entities that undertake one or more of the basic functions in industry itself, on consumers, and on the economy as a whole, the supply chain, namely generation, transmission, system and this important topic is reviewed in the fourth section. The control, distribution, and supply services to users of electricity. main lessons learned from designing and implementing these The supply services function encompasses the sale of electricity reforms are summarized in the final section. procured on the wholesale electricity market to electricity users and the associated customer services of billing, collection, and maintenance. These principles require that enterprises pay 4. Privatization of the unbundled electricity generators and taxes and market-based interest rates, earn commercially distributors under dispersed ownership, because competition is competitive returns on equity capital, and have the autonomy unlikely to develop properly between entities that are under to manage their own budgets, borrowing, procurement, and common ownership--whether state or private. In developing labor employment. countries, furthermore, private investors and operators are expected to bring in financial resources and technical and 2. Introduction of competition in order to improve sector managerial expertise that will rectify the prevailing low performance in terms of efficiency, customer responsiveness, standard of electricity supply by state-owned power utilities. innovation, and viability. Competition can be developed in the generation- and supply-service segments but in most cases 5. Development of economic regulation of the power market is not feasible in the network segments (transmission, that is applied transparently by an agency that operates distribution, and system control) because these functions are independently from influence by government, electricity natural monopolies. Supply services to large electricity users suppliers, or consumers. In the wholesale market, the focus of is an intrinsically competitive segment because the cost of regulation is to prevent anticompetitive abuses of market competing for their business is small compared with the power. In the retail market, the focus of regulation should be potential profits. Supply services to all but large electricity on balancing the interests of suppliers with the interests of their 4 users, however, has usually been a monopoly in practice captive customers. because the profits per customer are too small to stimulate competition. Hence, this element of supply service has generally 6. Focusing of government's role on policy formation and been carried out by the entity that distributes electricity to these execution while giving up the roles of operator and investor users because both these functions serve the same market. The with divestiture of state ownership in generation and threshold level of customer demand at which the supply to meet distribution. it becomes competitive has been coming down, however, and full competition in the retail market has been introduced in The process of a full reform program therefore consists of England and Wales, Norway, and some parts of the United the following four main stages: (a) formation and approval States. A further consideration is that consumers must be able to of a power policy by government that provides the broad switch between suppliers at low cost--any arrangement in which guidelines for the reform program and the heavy political consumers have to remain with their original supplier gives commitment needed to sustain the reform process, followed market power back to the sellers even when the sellers have by the enactment of legislation necessary for implementing this only modest market shares. policy; (b) development of a transparent regulatory framework for the electricity market; (c) unbundling of the integrated 3. Restructuring of the electric power supply chain to enable structure of the power supply and establishing a market in the introduction of competition. This involves breaking up which electricity is traded at arm's length; and (d) divestiture ("unbundling") the incumbent power utility into multiple of the state's ownership at least in most of the electricity generators and distributors of power that trade with each other generation and distribution segments of the market. in a competitive wholesale power market. To prevent the acquisition of anticompetitive amounts of vertical market power Reform programs for electricity sectors must be built around by any generators or distributors, transmission, and system these basic elements, but the detailed design of each program control are placed with independent companies (or they may be should reflect the particular circumstances of the country and combined) with restrictions on ownership or on control (through its electricity sector. Hence, actual reform programs exhibit governance arrangements) of such companies by generators and a variety of designs, particularly in terms of market structure, distributors. Independent electricity suppliers should be allowed degree of private involvement, and sequencing of reform to compete with distributors for the custom of large users (this stages. could be delayed in those countries where distribution and supply systems are so dilapidated at the time of privatization that The variety of market structures can be categorized according new owners need a period of assured revenues to remedy the to increasing degree of competition, as follows (15). Model 1 worst deficiencies before having to compete for the business of (monopoly) has no competition at all, only monopoly at all their largest customers), and supply licenses can be granted to levels of the supply chain. A single monopolist produces and generators as well as to firms that specialize in energy trading. delivers electricity to the users. Model 2 (purchasing agency) allows a single buyer or purchasing agency to encourage terms by the market regulator. This form is appealing for competition between generators by choosing its sources of countries with small power systems and weak institutional electricity from a number of different electricity producers. The capacity. Such contracts provide for competition only at the agency on-sells electricity to distribution companies and large time of bidding for the right to secure such contracts and do power users without competition from other suppliers. Model 3 not allow competition to develop as trade takes place in the (wholesale competition) allows distribution companies to market. Moreover, a group of bilateral contracts will not purchase electricity directly from generators they choose, match total supply precisely with the constantly changing total transmit this electricity under open access arrangements over demand for electricity in the market, and hence, a means of the transmission system to their service area, and deliver it over balancing supply with demand at the margin needs to be their local grids to their customers, which brings competition included in this design. One solution is to establish a balancing into the wholesale supply market but not the retail power pool in which suppliers and buyers trade at spot prices to market. Model 4 (retail competition) allows all customers to balance their needs, whereas another is for one generator to choose their electricity supplier, which implies full retail undertake to act as the "swing producer." Beyond these competition, under open access for suppliers to the transmission designs lie power pools of ever-increasing complexity that and distribution systems. allow progressively more competition in the market (17). The best examples of power pools outside Europe and North Reform programs are designed to progress through these America are found in South America (Argentina, Brazil, Peru, 5 models, starting from model 1 and progressing through model and Bolivia). 2 or 3 until eventually reaching model 4. This progression reflects the basic sequence for a reform program, whereby Independent power producers (IPPs) are often the first private restructuring the supply industry and setting up the legal and investors in a power market dominated by state-owned power regulatory framework precedes the transfer of ownership of utilities, and their entry can help to launch the reform process power generation and distribution from the state to the private by showing the benefits of private investment and management. sector. Models 3 and 4 are important for countries contemplating IPPs can enter the wholesale power market under any of the reforming their power sectors because they offer plausible four models described above, but they need the benefit of alternatives to the traditional European state-owned, integrated long-term power purchase agreements (PPAs) backed by industry structure, and to the rate-of-return based, investor- suitable guarantees for raising long-term financing under the owned utility model developed in the United States. first two models, and also in the third model where competition is still nascent, because of the regulatory risks faced by such Many reform programs in developing countries focus on investments. When the wholesale market is fully competitive, moving from model 1 to model 3. The key decision is whether under model 3 or 4, IPPs can dispense with PPAs for a to go for model 2 or model 3. Adoption of model 2 in some substantial portion of their output, which they can sell directly developing countries has been justified largely as a transition to the market. IPPs may earn the right to enter into PPAs under stage to model 3 that is needed to allow time for the a competitive bidding process that gives transparency to the generation and distribution sectors to develop sufficiently for process and thus sustainability to the agreement (18). the operation of a competitive wholesale electricity market. The main risk with this model is that government can still Intermediate between models 1 and 2 is a model whereby the impose uncommercial practices on the market by manipulating private sector is brought in under a long-term concession as the single buyer. An additional risk is that government's operator, but not investor/owner, of the incumbent integrated commitment to full reform may weaken to avoid politically utility. The anticipated benefit is to turn round a poorly controversial consequences of introducing more privatization performing utility in conditions that are unlikely to attract and competition (16). private investors because of their perceptions of high country and sector risk. In other words, concessions are viewed as an The design of the new market for trading electricity depends option when competition under models 3 and 4 is not on the industry structure that is adopted and on the capability considered feasible. Concessioning of small utilities has been of the suppliers and purchasers in the market to handle trading tried in francophone West African countries with limited mechanisms. The simplest design is a series of bilateral success [as in Senegal (19,20)]. contracts between buyers and sellers, basically electricity distributors and generators, subject to the approval of contract PROGRESS WITH SECTOR REFORM models 2 and 3 have emerged as the main options for the developing countries that have yet to select one. Electricity-sector reform has been advocated by such international lending agencies as the World Bank (21), the In Eastern Europe and the countries of the former Soviet European Bank for Reconstruction and Development, and Union, power-sector reform has progressed slowly, although a the Inter-American Development Bank (22), as well as by such wide range of reform approaches is present in the region, organizations as the World Energy Council. To assess what including (a) reform limited to IPPs (Croatia, Slovakia), (b) has been achieved by which countries, these bodies have third-party access to the dominant utility's network for private carried out analyses of the steps taken by their member producers (Czech Republic), (c) restructuring with the intention nations. At the same time, the private sector has a great of major divestiture of state ownership (Poland, Russian interest in knowing what is happening on a comparative basis, Federation, and Ukraine), and (d) opening up the power so that investment decisions can be put into proper context, market to new entry through concessions, including IPPs, and a number of publications have addressed this need for restructuring, and divestiture (Hungary, Kazakhstan, Georgia, information. and Moldova). This diversity of approaches indicates that it is too early to conclude definitively that the region is adopting a From a global perspective, the movement to reform power common reform model. However, because other countries in 6 sectors appears to be sweeping across the developing world the region (Bulgaria, Estonia, Latvia, and Romania) are also at a rate similar to that in the industrialized world. Many considering or have started on reforms that match model 3, developing countries have already started their reforms, and a general preference for the South American approach may some of them have substantially completed them. Moreover, be emerging in this region (23). many more, possibly most, are either planning or seriously considering similar reforms. This phenomenon can be Africa and the Middle East have lagged behind other regions explained by the presence of a strong demonstration effect. in implementing reforms to power sectors, except for the The pioneering reforms to power sectors in Chile, England concessioning of utility management of private operators and Wales, and Norway during the 1980s (which fall under (usually a foreign power utility) in some francophone countries. model 3) have motivated numerous industrialized and Algeria, Côte d'Ivoire, Egypt, Ghana, Kenya, Morocco, developing countries to follow them during the 1990s. A Senegal, and Tanzania have attracted one or more IPPs mixture of these two variants of model 3 (the power pool (model 2); Zambia has privatized a generation station and its design of the Chilean model, the independent transmission local transmission grid in the copper-belt area, whereas Togo and system operator of the England and Wales model) has has privatized its small power utility without restructuring under been widely adopted in South America (Argentina, Brazil, a 20-year concession. Africa appears to be catching up to Bolivia, Colombia, and Peru). Other countries have implemented other regions, however, because many more African countries variations on this model, particularly for the use of bilateral are considering reforms to their power sectors, mainly along contracts between producers and suppliers (Georgia, the lines of model 3. This approach would be supported by Hungary, and Moldova). Many other countries are in the the ongoing developments to form regional power pools in process of reforming their power sectors based on similar the southern, eastern, and western areas of the continent, approaches, some of whom (Ecuador, Armenia, Bulgaria, which would help compensate for the small size of the national and Romania) have completed the initial restructuring and power markets. Other African countries are in the process of regulatory steps but have yet to privatize most of their negotiating with IPPs and/or putting out some or all of their generation and distribution entities. distribution markets to long-term concessions. Also in the 1990s, the model of IPPs selling to a state-owned The two global studies on the extent of electricity-sector reform power utility (model 2) spread across Asia (China, India, refer to the sector as it was in 1998, so that the findings Indonesia, Malaysia, Nepal, Pakistan, Philippines, Thailand, reported below do not fully reflect the current state of affairs. Vietnam, and Jordan) and Central America and the A study by the World Energy Council (24) gave an assessment Caribbean (Guatemala, Honduras, Belize, Jamaica, Panama, for all its member countries (both developed and developing) and the Dominican Republic). Model 4 has been adopted in of the state of the energy industries. The basic headings (with England and Wales (where model 3 was transitional) and in some variation between fuels) were ownership and control, Norway, but not yet in any developing countries. Hence, privatization extent, entry barriers, pricing basis and regulation, international trade restrictions, and fiscal framework. These · "Has any of the existing state owned enterprise been indicators were combined into an overall indicator to give an privatized [including outright sale, voucher privatization or assessment of the degree of sector liberalization. joint ventures]?" Table 1 shows the number of countries (developed and developing) judged to have made substantial progress to TABLE 2 Number of countries taking key reform steps sector liberalization, to be underway to sector reform, or to in the power sector as of 1998a have made no steps toward sector reform. CORPORATE LAW REGULATOR IPPS RESTRUCTURE GENERATION DISTRIBUTION REFORM PRIVATIZATION PRIVATIZATION INDICATOR 51 38 33 46 40 24 21 2.06 TABLE 1 Number of countries achieving substantial (44%) (33%) (29%) (40%) (35%) (21%) (18%) (34%) power-sector liberalization by 1998a a Energy Sector Management Assistance Programme data. IPP, Independent REGION SUBSTANTIAL LIBERALIZATION NO power producers LIBERALIZATION PLANNED/UNDER WAY LIBERALIZATION Western Europe 5 12 2 Cent. & EasternEurope/CISb 4 8 15 For each of the 115 countries analyzed, the maximum reform Africa/Middle East 0 17 46 score was six (all steps taken) and the minimum was zero (no Asia/Australia 2 12 13 steps taken). Table 2 shows the number of countries that had 7 South America 4 3 5 North America 0 3 0 taken key reform steps. The privatization of assets was least Total 15 55 81 common, with approximately 20% of countries having undertaken some action in that direction; the most common a World Energy Council criteria. Excluding Central America and the was corporatization and commercialization of the state utility, Caribbean, for which full details were not given. b with over 40% of countries having taken this step. CIS, Commonwealth of Independent States. The overall indicator shows that for the developing countries As shown in Table 1, only 15 countries actually carried out surveyed, on the average one third of these reform steps had substantial reforms of previously state-owned energy industries. been undertaken. The method of scoring tended to exaggerate Furthermore, analysis of the individual country data reveals the extent of reform because any action toward privatization that about half of these were high-income, industrialized (however small a share of all assets had been sold) counted as a countries. Some 55 countries had liberalization under way or "success." The study also allowed investigation of whether there planned, and of these about one third were in the high- were important differences between countries, and whether any income, industrialized group. Finally, some 81 countries had such differences were systematically related to features of the made no move toward sector liberalization--many of these countries involved. The data on the overall reform indicators were less-developed countries. were first analyzed by grouping countries into the World Bank's regions. The results are shown in Table 3 (see page 8). An Energy Sector Management Assistance Programme (ESMAP) publication (25) focused entirely on the non-industrialized Table 3 shows the great unevenness in reform effort between countries. The questionnaire asked the following questions, regions. In the Latin America/Caribbean region, almost three which were to be answered "yes" or "no." quarters of the reform steps had been taken, whereas in the Africa/Middle East/North Africa region, only one sixth of the ·"Has the utility been commercialized and corporatized?" steps had been taken. The actual privatization of assets and the ·"Has an `Energy Law' been completely passed by introduction of independent power producers were similarly Parliament [a law which would permit the creation of a unevenly spread, so that the key element of the reform program, sector that could be unbundled and/or privatized in part or which was the introduction of the private sector with its finance, in whole]?" know-how, and management skills, had hardly started in a large ·"Has a regulatory body started work [a body that is number of developing countries. This inequality between countries separate from the utility and from the Ministry]?" is shown clearly by the following statistics: Out of the 115 ·"Is there any private sector investment on greenfield countries surveyed, 42 had taken no reform steps and 15 had sites in operation, or under construction?" taken only one step. Only 10 countries had taken five steps and ·"Has the core state owned utility been only 12 had taken all six steps, so that the Gini coefficient of restructured/separated?" reform inequality was as high as 55%. TABLE 3 Number of countries having taken key reform steps by region as of 1998a REGION (NO. COUNTRIES) KEY STEP AFR(48) EAP (9) ECA (27) LCC (18) MNA (8) SAR (5) Corporate 15 (31%) 4 (44%) 17 (63%) 11 (61%) 2 (25%) 2 (40%) Law 7 (15%) 3 (33%) 11 (41%) 14 (78%) 1 (13%) 2 (40%) Regulator 4 (8%) 1 (11%) 11 (41%) 15 (83%) 0 (0%) 2 (40%) IPPs 9 (19%) 7 (78%) 9 (33%) 15 (83%) 1 (13%) 5 (100%) Restructuring 4 (8%) 4 (44%) 14 (52%) 13 (72%) 3 (38%) 2 (40%) Generation privatization 2 (4%) 2 (22%) 10 (37%) 7 (39%) 1 (13%) 2 (40%) Distribution privatization 2 (4%) 1 (11%) 8 (30%) 8 (44%) 1 (13%) 1 (20%) Reform indicator 0.88 (15%) 2.44 (41%) 2.70 (45%) 4.28 (71%) 1.00 (17%) 3.00 (5%) aEnergy Sector Management Assistance Program data. AFR, Africa; EAP, East Asia and Pacific; ECA, Europe and Central Asia; LCC, Latin America and the Caribbean; MNA, Middle East and North Africa; SAR, South Asia; IPP, independent power producers. In the summer of 2000, the World Bank carried out a further Level 3: Law passed accounting for full-scale restructuring of 8 analysis of power-sector reform for 116 developing countries. the industry, including vertical unbundling through accounting In 17, industrial enterprises did have a choice of power separation, setting up a regulator; some tariff reform and supplier; in 37, there was a regulator who was judged to be improvements in revenue collection; some private involvement. operating in an objective, transparent, and nondiscriminatory manner in order to promote competition; and in 27, private Level 4: Law for industry restructuring passed with ownership and financing was judged to play a dominant role separation of the industry into generation, transmission, in the power sector. These figures suggest that there had been and distribution, and setting up of a regulator with rules for some progress in the entry of the private sector, but that true cost-effective tariff-setting formulated and implemented; competition had not been widely established. These findings arrangements for network access (negotiated access, have been confirmed also for the Mediterranean region (26), single-buyer model) developed; substantial private-sector including countries not covered by the ESMAP study. involvement in distribution and/or generation. The European Bank for Reconstruction and Development (27) Level 4+: Business separated vertically into generation, defined the levels of sector transition according to more transmission, and distribution; an independent regulator with complex criteria, which were as follows: full power to set cost-reflective effective tariffs; large-scale private-sector involvement; institutional development covering Level 1: Power sector operated as a government arrangements for network access and full competition in department; political interference in running the industry; generation. few commercial freedoms or pressures; average prices below costs, with external and implicit subsidy and cross subsidy; Using these criteria, by the year 2000, of the 26 countries very little institutional reform, with monolithic structure with in the survey, one scored a 4, eight scored 3, thirteen scored no separation of different parts of the business. 2, and four scored 1. None had reached the 4+ level. Thus, similar pictures emerge from the World Bank and the World Level 2: Power company distanced from government, Energy Council studies, i.e., that in many countries of Eastern for example a joint-stock company, although still political Europe and Central Asia, there is much to be done before interference; some attempts to harden budget constraint the sector can be considered to be reformed extensively. but management incentives for efficient performance weak; some degree of subsidy and cross subsidy; little institutional For developed countries, the most elaborate scorecard for reform; monolithic structure with no separation of different measuring power-sector reform is the RED (Retail Energy parts of the business; minimal, if any, private-sector Deregulation) index (28), which gives an overall reform score investment. in the year 2000 for each of 51 states/jurisdictions within the United States based on 18 attributes. On a scale of 0­100 (maximum reform), the most advanced state scored 65, and six states scored 0, showing no reform. This also confirms supply and less need for finance to expand the sector: With the great variability, even within the United States, of the less pressure for growth, there would be less pressure for reform. present state of sector reform. In addition, six "dummy" variables are constructed. Each The great variability between countries leads naturally to region has a dummy variable that (a) takes a value of unity if considerations of whether there are patterns in the extent of the country in question is in that region and (b) takes a value reform. Individual country reform scores from the ESMAP study of zero otherwise. The dummy variables measure any common (25) were correlated with economic variables that might be tendency for countries in a given region to have a higher (or expected to show a relationship with the degree of reform lower) reform score than the average of countries from all achieved by 1998. Data were collected on several economic other regions. variables: gross domestic product (GDP) per capita in 1997 US dollars; a country policy and institutional assessment for The cross-section regression in Table 4 shows that the risk 1998; country risk for 1998; aid/GDP in 1997; commercial indicator is significantly correlated with the level of reform-- energy use per capita in kilograms of oil equivalent for 1996; the less-risky countries have higher reform scores. The World and annual growth rate of commercial energy use per capita Bank's policy assessment indicator also is significant--the between 1980 and 1996. better the policy assessment, the more reform steps are likely to have been taken. In addition to these effects, there are 9 The country policy and institutional assessment is based on effects for three regions relative to the other three that show 20 indicators, which focus on macroeconomic management nosignificant differences between themselves once the level of and sustainability of reforms, policies for sustainable and risk and policy have been considered. Countries in the Latin equitable growth, policies for reducing inequalities, and America/Caribbean region have taken about one more reform public-sector management. The hypothesis would be that step for the same risk and policy assessment, whereas this indicator should be positively correlated with the countries in Africa have taken one step less, and countries in reform score because all components are conducive to the Middle East/North Africa region have taken two steps less more active reform. than the group of other countries. The aid dependency variable does not show a significant effect independent of the The risk indicator is based on a weighted average of nine other variables, which may reflect the fact that it is directly indices, of which political risk and economic performance each or indirectly incorporated into the other variables. The lack of account for 25% of the weighting. Reform is hypothesized to be a significant incremental effect of GDP per capita is striking. positively correlated with this indicator, i.e., less reform occurs A plot of the data tends to suggest that countries that have in countries that are assessed as more risky (low-risk taken a large number of steps include both high- and low- indicator score). income levels [e.g., Argentina at $8950 (US) versus Bolivia at $970 (US)], so that there is no strong pattern ranked by The GDP per-capita variable measures the general level of income level even among the most reforming countries. economic development, and it is hypothesized that this would be positively correlated with the reform indicator. TABLE 4 Multiple regression for power-sector reform The ratio of aid to GDP gives a measure of the dependence indicatora of the economy on foreign aid. It is hypothesized that the VARIABLE COEFFICIENT STANDARD t-STATISTIC PROBABILITY higher the dependency ratio, the less the country will have ERROR reformed. Constant -0.476 0.815 -0.583 0.560 Risk 0.035 0.019 1.859 0.065 The energy use per capita is an attempt to measure the Policy 0.630 0.309 2.034 0.044 1.019 0.479 2.126 0.035 extent to which the economy has become energy using--the LCC MNA -2.140 0.658 -3.252 0.001 higher the value, the more likely it would be to reform. The AFRICA -1.146 0.374 -3.063 0.002 growth in energy use measures the pressure on the energy R-squared 0.473 Mean dependent 2.110 sector. Rapid growth of demand should require more variable efficient supply and hence reform, whereas a decline in aWorld Bank calculations based on Energy Sector Management Assistance Program reform scores. LCC, Latin America and the Caribbean; MNA, use (as happened in many countries) might indicate excess Middle East and North Africa. In addition to the total number of steps taken, the design of Using the same data, an index of the optimal sequencing of reform can be assessed in terms of the combinations of steps reforms carried out was calculated. For the given number of taken or by the sequence in which the steps were taken. Both reform steps taken, a perfect score of 0 would be achieved if approaches assign an ideal order to the steps. Each step they were taken in the optimal sequence, whereas the maximum actually taken can be seen as less effective if the earlier steps score of 100% would occur if the steps were those of the in this ideal sequence have not been taken. optimal sequence taken in reverse order--for example, if only two steps were taken, then taking step g above first, followed The ESMAP study (22) had used the following reference by step f, would be the worst possible sequence and would sequence for its six steps: corporatization, restructuring, law, score 100%. Table 5 gives sequencing scores for a group of regulator, IPP entry, and divestiture. Hence, for example, a countries. It shows that although many countries took the steps country that by 1998 had corporatized and restructured would in a sequence closely matching the "optimal" sequence, there have taken two steps in an ideal combination, whereas a were several whose sequence was far from optimal. country that had corporatized and had some IPPs would not have the ideal combination because it did not possess a law or a regulator, nor had it unbundled--it would have just one TABLE 5 Index of optimal sequencing of power-sector step of the reference. Using this criteria for the reference reforms taken to datea 10 combinations, the survey showed that 63 countries had taken COUNTRY INDEX OF SEQUENCING NO. STEPS no steps in the reference combination, 18 had taken one step; (% WORST POSSIBLE) TAKEN 9, two steps; 5, three steps; 5, four steps; 2, five steps; and 12, Albania 3 4 all six steps (necessarily the optimal combination). This simple Argentina 5 7 calculation, which is dependent on the definition of the reference Bolivia 0 7 combination, nevertheless does suggest that the amount Brazil 46 7 of well-structured reform is even less than the basic statistics for Colombia 48 5 reform steps taken would indicate. In fact, further experience Cote d'Ivoire 55 6 with reform suggested that an alternative sequence to the Georgia 61 7 reference sequence would be better, as discussed below. Hungary 2 7 India (Orissa) 2 6 75 7 In a later study, for a subset of countries, the World Bank Kazakhstan Mali 30 5 collected data on the dates at which various steps were taken Panama 9 7 in order to calculate the duration of the reform process and Thailand 39 3 the extent to which the temporal sequence of steps taken was Turkey 34 6 optimal. In this study, the optimal temporal sequence was (a) Ukraine 34 5 launch a privatization/liberalization program, (b) enact an a From R. W. Bacon & J. Besant-Jones, unpublished data. electricity law permitting unbundling and divestiture, (c) establish an independent regulatory authority, (d) approve a new power market structure, (e) unbundle the power utility, (f) The outcome of the reform process can also be measured by privatize or close on a concession for some private distribution, looking in finer detail at the steps taken. Several countries and (g) privatize some generation. The entry of IPPs (under have unbundled and privatized generation. Here, one issue is terms that did not hinder other necessary reform steps) was whether this was done in such a way as to avoid abuse of seen as a step that might happen at any time in the sequence market power from the creation of just one or two dominant without being flawed if other steps had not been taken. A firms (29,30), which was a problem with the original breakup subset of countries surveyed had by the year 2000 taken all of the sector in England and Wales (31,32,34,35,) and in these steps. The span of time required to go from the initial Chile (36). A simple measure of the degree of potential market formal launching of a program for reform of the sector to power is given by the Herfindahl-Hirschman Index (HHI).2 carrying out all these steps varied markedly between countries, This has been calculated for a series of countries (regions) that as the following dates show: Argentina, 1989­1992; Bolivia, have undertaken extensive restructuring and privatization. 1993­1995; Brazil, 1990­1999; Georgia, 1994­1999; 2 Ghana, 1994­1999; Hungary, 1994­1995; Kazakhstan, The HHI is defined as Si2 , where Si is the share of the i'th firm's capacity in the market. The index varies between unity for a monopoly and 0 for perfect competition (a very large number 1995­1999; and Panama, 1995­1999. of equal-size firms). Two relationships are of importance for interpreting the HHI. First, the value given by its reciprocal is equal to the number of identical-size firms that would have the same concentration ratio. Second, the weighted average price over cost margin (Lerner index) given by Si[p-ci]/p under conditions of Cournot competition (competition by quantity) is equal to the HHI divided by the price elasticity of demand in that market. Hence, the HHI can be taken as an indicator of how much the price could be raised above the marginal costs of production where there is no regulation to control prices. TABLE 6 Index of market concentration for developing countries is provided by the World Bank's Private power-sector generationa Participation in Infrastructure database. COUNTRY/REGION NO. FIRMS SHARE OF LARGEST HHI EQUIVALENT NO. FIRM (%) EQUAL-SIZE FIRMS Table 7 gives data on power projects in developing countries with private sector participation that came to closure in the years Argentina 38 14 0.06 16.7 California 40 23 0.11 9.1 between 1990 and 1999. The total private-sector involvement in Australian NEW 11 18 0.12 8.3 the power sector over these 9 years was approximately $150 Colombia 26 24 0.14 7.1 billion (US)--showing that very large sums could be attracted Brazil 14 25 0.15 6.7 to the sector. More was invested in greenfield projects than in England and Wales 32 28 0.16 6.3 divestiture, but the amount per project was almost identical. The New England 16 32 0.18 5.6 pattern over time showed a rapid increase in the number of Bolivia 6 26 0.19 5.2 private-sector projects until 1993, after which it kept roughly Hungary 10 27 0.19 5.3 constant until 1998, when there was a major collapse. Regionally Peru (SICN) 8 35 0.23 4.3 there were enormous differences. Of all this investment, 40% Sweden 8 52 0.32 3.1 went to the Latin America/Caribbean region (both greenfield and Chile (SING) 4 43 0.33 3.0 divestiture) and 36% went to the East Asia/Pacific region (mainly N. Ireland 4 48 0.33 3.0 greenfield). Less than 2% went to Africa, and less than 4% went 11 Spain 8 46 0.34 2.9 Alberta 12 55 0.38 2.6 to the Middle East/North Africa region. Chile (SIC) 4 60 0.43 2.3 New Zealand 6 68 0.53 1.9 TABLE 7 Electricity projects with private participation Czech Rep. 6 75 0.60 1.7 in developing countries that reached financial Queensland 2 76 0.64 1.6 Portugal 3 93 0.86 1.2 closure 1990­1999 a From R. Lamech, unpublished data. HHI, Herfindahl-Hirschman Index; CLOSURE DATE NUMBER DIVESTITURE GREENFIELD PROJECT EXPENDITURE EXPENDITURE SICN, Central Northern System; SING, Northern Interconnected System; SIC, Central Interconnected System. 1990 7 803 409 1991 3 147 863 1992 25 3468 2987 Table 6 shows that even in some large, high-income industrialized 1993 105 2747 5955 countries, where it would have been possible to create several 1994 72 3936 9588 approximately equal-size private generators, the structure 1995 94 5190 13069 chosen has created considerable market power. For example, in 1996 109 11096 17541 Spain, although there are eight firms, the largest controls 46% 1997 117 21428 16349 of the total generating capacity, so that the structure has the 1998 57 11103 8478 same market power (and ability to raise prices above costs) as if 1999 56 4904 7421 there had been three equal-size firms. Argentina, which deliber- Total 645 64822 82660 ately designed the reform so that no firm could have more than aIn millions of dollars (US). From World Bank PPI database, 2000. 15% of the market, has the lowest HHI, and Colombia and Brazil also have low values. The Czech Republic and Chile (where one firm has 60% of its market) have very high values of In addition to these comparative studies of the steps taken to the HHI. Two interesting cases are Bolivia and Peru--small reform power sectors in various countries, there have been a countries, with relatively few generating plants to be privatized-- number of studies describing in more detail what happened in where both managed to avoid creating the very high levels of each case. An early review of the state of reform was given in market power found elsewhere. Moscote et al. ( 37) for all the Latin American countries. Several African countries are covered in Chiwaya et al. ( 38), Argentina Because one major reason for the reform movement has been in Bastos & Abdala ( 39), Israel in Czamanski (40 ), European to attract private finance into the sector (either for greenfield Union countries in Lauriol ( 41), and Latin America and the investment or for the purchase of existing state-owned assets), Caribbean in Suding (42 ). Power sector regulation in several an important measure of progress of reform is the amount of countries, both developing and developed, has been analyzed private finance that has entered the sector. A useful summary for by a series of authors (43 ). RESULTS TO DATE OF SECTOR REFORM The four companies showed substantial improvements in performance according to all these indicators. These The previous section emphasized that many paths to reforming improvements show the benefit of having private management power sectors have been taken by developing countries, both focus on commercial performance, which has been a major in the number of steps taken and in the sequence of steps. weakness of state-owned utilities. Some countries have embraced a program of majority privatization of the sector to introduce competitive elements Some countries have obtained substantial proceeds from the wherever possible. Others have made tentative progress along privatization of their distribution entities, notably Brazil [more the IPP route. Many others have done little or nothing. than $20 billion (US)]. A comparison of these proceeds for Nevertheless, patterns in country choices can be identified that 12 developing countries is given in Table 9, based on the help explain the momentum developed for reform throughout values of the winning bids for the distribution entities. To the developing world. Moreover, a track record is beginning control for differences in scale and timing of the transactions, to show for the main approaches to reform that provides an the comparison is made on the basis of enterprise value per interesting comparison of their relative performance and some customer served on a countrywide basis, where enterprise lessons for countries that have yet to carry out or complete value is calculated as the total market value of the enterprise their reforms. (debt plus equity) based on the value of the winning bid for 12 a defined proportion of shares in the enterprise, and where The improvement in efficiency after privatization of four South the countrywide value is the sum of the current values of these American distribution companies is summarized in Table 8. enterprise values. The results show a wide range of values, These improvements are measured in terms of the change in from over $1300 (US) in Colombia, Brazil and Panama, to performance between the date of privatization and 1998. The around $300 (US) or less in India, Georgia, and Bolivia following range of performance indicators were selected for this (M. Hoskote, A. Marghub & S. Ostrover, unpublished data). purpose: (a) energy sales (gigawatt hours/year)--positive if This range reflects differences in investors' perceptions of the amount increased; (b) energy losses (percentage)--positive if rewards and risks under the business environment in each level declined; (c) employees--positive if amount declined; (d) country, notably in the stability and transparency of the regulatory customers per employee--positive if number increased; (e) net process, the administration of the process for selling the shares receivables (days)--positive if level declined; and (f) provisions in the distribution enterprises, stability of macroeconomic for bad debts (percentage of sales)--positive if level declined. conditions, and potential for fitting the target enterprise into a broader business development strategy. TABLE 8 Improvement in performance of four South American electricity distribution companies from the TABLE 9 Privatization proceeds from sale of time of privatization until 1998a electricity distribution entitiesa DETERMINANT PERUb ARGENTINAc ARGENTINAd CHILEe COUNTRY ENTERPRISE VALUE PER CUSTOMER Year privatized 1994 1992 1992 1987 Colombia $1,681 Energy sales (GWh/year) +19% +79% +82% +26% Brazil $1,369 Energy losses (%) -50% -68% -63% -70% Panama $1,334 No. employees -43% -60% -63% -9% El Salvador $872 Customers/employee +135% +180% +215% +37% Dominican Republic $860 Net receivables (days) -27% -38% n.a. -68% Guatemala $773 Provisions for bad debts (% sales) -65% -35% n.a. -88% Argentina $763 Peru $637 Hungary $527 aChange in 1998 measured in terms of performance relative to the year Bolivia $364 of privatization. From company annual reports and websites. Georgia $304 bLuz Del Sur, distribution company. India (Orissa) $168 cEdesur, distribution company. dEdenor, distribution company. eChilectra, a distribution company. From M. Hoskote, A. Marghub & S. Ostrover, unpublished data. In the case of long-term power purchase agreements signed TABLE 10 Risk exposure to the impact of IPP costs in with IPPs by the incumbent power utility in a developing country, four Asian countriesa the first PPAs usually carry substantial government guarantees DETERMINANT INDONESIA MALAYSIA PHILIPPINES THAILAND for the performance of the utility in keeping to its obligations. These PPAs usually run for around 25 years in order to support IPP capacity in operation in 2329 7121 3676 2419 the financing of the heavy investment in fixed assets by the mid-2000 (MW)b IPPs while keeping the power sales price at an affordable level Exchange rate exposure through Low Low High High for the purchaser during the initial period of the power purchase origin of fuel supply agreement, which because of its need to cover repayment of Exposure to exchange rate through High Low High Low the project debt is when the price is highest. currency of wholesale tariff Exposure to exchange rate through High Low High Low The cumulative obligations of a utility to purchase power foreign debt for project financing under many of these agreements may expose it to serious Exposure to market risk through High High High Low proportion of domestic power financial risks, as occurred in many Asian countries as a result needs supplied by IPPs of the 1998 financial crisis, when their currencies devalued Exposure to off-taker payment High Low High Low but retail power tariffs were not allowed to rise (44 ). The problems through margin of greater these risks, the greater the obstacles they pose to 13 retail tariffs over wholesale prices restructuring the power utility as part of more radical aRisk exposure assessments adapted from Gray & Schuster (45). IPP, liberalization of the power market, because the prices that independent power producers. emerge from a liberalized wholesale power market undercut bFrom Reference (55) the sales prices under these off-take contracts. When this happens, the high-priced (in local currency terms) power purchase agreements become stranded costs that have to be The results show a wide difference in risk exposure just among absorbed under the restructuring of power supply, which can these four countries. Philippines has the greatest overall create a major obstacle to this reform. exposure, with a high rating for all five exposure indicators, which is creating enormous problems (45). Indonesia also has The risk exposure of utilities that are off-takers for many a high overall exposure, with a high rating for four indicators, contracts with IPPs depends on how these risks are structured. whereas Thailand has a moderately low overall exposure, with In some cases, the utilities have taken on substantial risks, a high rating for two indicators, and Malaysia has a low whereas in other cases the utilities are much less exposed to overall exposure, with a high rating for only one indicator. risks beyond their control. This difference is illustrated in Table 10 for four Asian countries (Indonesia, Malaysia, Philippines, Poland, the Dominican Republic, and Pakistan have problems, and Thailand) that have followed the IPP route to reform. The similar to that of the Philippines, with a single-buyer approach following five types of exposure are assessed: (a) exchange to contracting with IPPs. In Poland, the transmission company rate exposure through origin of fuel supply--high if the fuel is took on long-term PPAs with all the generating companies imported; (b) exposure to exchange rate through currency of formed from restructuring the sector but at prices that were wholesale tariff--high if the currency is denominated in US later undercut by prices realized in the new competitive dollars or another hard currency; (c) exposure to exchange wholesale power market (46). In the Dominican Republic (47) rate through foreign debt for project financing--high if the and Pakistan (48), the problem stemmed from arrears in foreign debt made up more than 50% of project financing; (d) payments by the state-owned utility to the IPPs caused by low exposure to market risk through proportion of domestic power retail tariffs and low collection of payments from power needs supplied by IPPs--high if this proportion is over 50%; users. To keep the utility's risk exposure within manageable and (e) exposure to off-taker payment problems through margin proportions, only a few such PPAs should be signed before of retail tariffs over wholesale prices--high if this margin is the power market is reformed. Where the pressure to sign less than 3 cents (US) per kilowatt hour for covering the costs PPAs is caused by the need to reduce or eliminate a costly of transmission, distribution, customer services, technical losses shortage of power-supply capacity to meet demand, an in the power system, and nontechnical losses caused by theft alternative to a long-term PPA is a short-to-medium-term PPA of power and late or nonpayment of bills by customers. with an IPP to supply power from barge-mounted or skid- mounted generating units that can be installed in fewer than 6 months from closure, and that require far less investment governments must sustain an attractive business environment than needed for a plant installed under long-term PPAs. Of and sound sector regulation to attract the required amount course, the price of power under this alternative tends to be of investment in power capacity on competitive terms. higher than under a long-term PPA because of higher fuel consumption by the units and shorter term for amortization of Domestic capital markets are too undeveloped to replace capital expenditures, but the benefit of quick additions to foreign finance or to provide a market assessment of supply can be an advantageous trade-off for the host country. performance by power suppliers and regulators. Hence, This approach has been taken in Bangladesh (49), Nigeria, developing countries should avoid giving perceptions of the Philippines, Jamaica, the Dominican Republic, and excessive risk in their power sectors to foreign investors in the Guatemala. global competition for finance. For example, protection against major uncertainty in the regulation of tariffs and licensing is needed to attract private investors during the period LESSONS FROM SECTOR REFORMS following reforms until a good record has been established by the government and the new regulator. This requires that Experience with designing and implementing reforms to power regulatory powers over electricity prices, for example, be sectors yields several lessons for power-sector reform (50). limited to applying rules and regulations laid down in 14 secondary legislation for a specific period following privatiza- Radical restructuring of an integrated power supply chain tion of distribution and supply. This approach can be of functions is feasible--generation, transmission, and implemented without undermining the long-term regulatory distribution can be separated from one another even in power framework by granting vesting contracts to the new distribution sectors that did not adopt this structure from an early stage of companies for a limited period (approximately 5 years), during their development. There is sufficient track record to provide which certain regulated variables are specified. assurance that restructuring is possible while still ensuring proper coordination among these power-supply functions and A bid-based competitive power pool based on spot pricing is maintaining security of supply to power users. too complex to operate and too difficult to monitor for abuse of market power for all but the most advanced developing Private financing of power investments is feasible in a sound countries--even California is having well-publicized problems business environment, and commercial lenders are willing to with this approach (51). Moreover, the small size of the power provide medium-term financing (10­15 years) for investments market in many developing countries (about 100 countries in well-functioning reformed power sectors that are establishing have power markets of under 1000 MW) would limit the a good track record of adhering to sound regulatory principles number of viable participants that can be formed from (except during a global or regional macroeconomic crisis). unbundling a utility to fewer than is needed to sustain Conversely, in countries with little record of sound regulation competition in the market (52). There are simpler approaches and fair dealing for businesses, commercial lenders are to managing the wholesale power market in these countries, conspicuous by their absence or by their willingness to lend based on a well-designed set of market rules according to only on short maturities (under 3 years). The governments of production costs (for example, the cost-based bidding these countries have to carry an unduly high proportion of approach used by South American countries). By allowing investment risks through performance and payment guarantees, competition for market share, these approaches can give aided by the presence of multilateral financial participation, in incentives to producers to reduce costs (Argentina, Bolivia, order to attract large investments to their power sectors from Chile, Peru, and Poland). the private sector. For developing countries with fast-growing power demands The competition to developing countries from investment that exceed the available supply capacity for the foreseeable opportunities in industrialized countries has been stiffening future, the persistence of large supply shortages also rules during the past few years and magnifies the task of mobilizing out the possibility of competitive power pools because the the billions of dollars needed every year to finance new development of competition requires adequate supply capacity power-supply capacity in developing countries. This issue is to meet all segments (base, peak, and shoulder) of the load as relevant to a country whose power sector is under private on the power system. ownership as to one that is under state ownership. Hence, Competition for the right to enter the power market on Power-sector reform can yield huge productivity gains, contractual or regulated terms plays an important role in particularly through dynamic efficiency gains under competitive developing countries, notwithstanding the limited scope for pressures. However, regulators have difficulty in making power sustaining competitive power pools. For example, provided suppliers pass on some of their productivity gains through that sufficient interest can be attracted from bidders, lower retail power prices to franchised electricity consumers in governments and utilities can obtain better terms for the host noncompetitive retail market segments (in Argentina and country under competitive bidding for proposals from IPPs Chile, for example). This is because regulators also have to than under noncompetitive negotiated deals. Likewise, a avoid creating or adding to substantial uncertainty about transparent and soundly structured process for the sale of future revenues for private investors in power-supply facilities. stakes in power entities will yield the best terms for the long- term efficiency of the power sector. Governments and regulators must expect to face unanticipated challenges when radical reforms are proposed or introduced The sequencing of reforms is crucial to their long-term into their power sectors. Such challenges may come for groups sustainability. First, the legal and regulatory framework should that lose from the reforms, such as workers laid off under be in place before privatization of the restructured power productivity improvements, power users and fuel suppliers that supplier. Second, major restructuring should precede the have benefited from subsidies, and local manufacturers of creation of private ownership rights to avoid problems with plant and equipment preferred by the power utility before the 15 stranded assets. Third, the scope for introducing competition reforms but not competitive with other technologies preferred to the wholesale power market should be incorporated into by the new private owners. Private owners may even pull out the initial structural reforms to the power market, rather than of a power market if they cannot earn competitive returns, as relying only on later regulatory interventions to reduce the in Kazakhstan (53). market power of the largest generating companies. Fourth, the incumbent utility should not sign many long-term power Completion of a reform program is not the end of the process off-take agreements with IPPs before it is restructured and the of change in a liberalized power market. For example, once regulatory framework for a liberalized power market is in most of the power-supply industry has been passed into place. Fifth, where cash collections fall far short of the private ownership and is exposed to competitive forces, the revenues that should be collected by the incumbent power private owners may carry out further restructuring with moves utility from power consumers--regrettably, a situation that to recombine some generation capacity with some distribution exists in many developing countries--the priority for the capacity to reduce market risks (England and Wales), or they privatization strategy should be to improve this performance may sell their stakes to other private parties under realignment by privatizing the distribution and supply functions first. This of their investment strategies, as in Brazil (54). These would help attract potential bidders for the upstream tendencies require careful antimonopoly regulation to generation facilities by signaling that the distributors and maintain competitive pressures on power suppliers. Another suppliers will become creditworthy buyers of power from example is when unexpectedly large profits by the new private the generators. producers and suppliers arouse public hostility to the reforms and provoke the regulator into making unscheduled price The timing of reform is also critical, particularly relative to reviews or the government into considering a windfall tax on the electoral cycle, for the privatization of electricity generators these profits (as happened in England and Wales). A third and distributors, and for an unpopular increase in electricity example is the unbearable upward pressure of retail tariffs tariffs needed to remove major subsidies. The success of a caused by unanticipated large currency devaluations that privatization program often depends on divesting most of can lead to demands from the utility for reductions in the the state's ownership before the government faces the next off-take prices under PPAs with IPPs, as described in the election, and this can force a compromise with long-term previous section in the case of some Asian countries following efficiency objectives for the sector (as happened in England their financial crisis in 1998. However, even though such and Wales). occurrences have been unpopular, there has been no move to reverse or undo the main steps of any country's liberalization program. 16 LITERATURE CITED Electric Utilities, ed. Z. Zaccour, 203-27. Kluwer Academic Publishers, Boston. 342pp. 14. Patterson R. and Cornwall C. 2000. Light-handed regulation of electricity networks in New Zealand: The failure of the New Zealand model. Journal of Network Industries 1: 89-99 15. Hunt S. and Shuttleworth G. 1996. Competition and Choice in Electricity. Wiley, Chichester, UK. 237 pp. 16. Lovei L. 2001. The Single-Buyer Model: A Dangerous Path toward Competitive Electricity Markets. Viewpoint 225. The 1. Bacon R. W. 1995. Privatization and Reform in the Global World Bank, Washington D.C. 4 pp Electricity Supply Industry. In Annual Review: Energy and the 17. Barker J. Jr., Tenenbaum B and Woolf F. 1997. Environment. 20: 119-43 Governance and Regulation of Power Pools and System 2. World Energy Council. 1998. The Benefits and Operators: An International Comparison. World Bank Deficiencies of Energy Sector Liberalisation, Volume I. World Technical Paper 382. The World Bank, Washington D.C. Energy Council, London. 121 pp.. 80 pp. 3. Patterson, W. 1999. Transforming Electricity. Earthscan, 18. World Bank, US Agency for International Development. 17 London. 203 pp. 1994. Submission and Evaluation of Proposals for Private 4. International Energy Agency. 1999. Electricity Market Power Generation Projects in Developing Countries. The Reform: An IEA Handbook. OECD/IEA, Paris. 106 pp. World Bank, Washington D.C. 97 pp. 5. Kosmo M. 1987. Money to Burn? The High Costs of Energy 19. Financial Times. 2000. Senegal's new broom sweeps out Subsidies. World Resources Institute, New York. 68 pp. Hyrdo Quebec and Elyo. Africa Energy 32: 4-5 6. International Energy Agency. 1999. World Energy Outlook: 20. Financial Times. 2000. Back to the market for Senelec. Looking at Energy Subsidies: Getting the Prices Right. Africa Energy 33 (Dec.):5 OECD/IEA, Paris. 220 pp. 21. World Bank. 1993. The World Bank's Role in the Electric 7. The World Bank. 1995. Bureaucrats in Business: The Power Sector. Policies for Effective Institutional, Regulatory Economics and Politics of Government Ownership. Oxford and Financial Reform. The World Bank, Washington D.C. University Press, Oxford. 346 pp. 84 pp. 8. Bradburd R. and Schulz E. 1997. Why Things Have to Get 22. Inter-American Development Bank. 2000. Energy Sector Worse Before They Can Get Better: Institutions, Interest Strategy. Sector Policy and Strategy Paper Series.: The Inter- groups, Decline and Reform. Research Paper Series of the American Development Bank, Washington D.C. 34pp. Department of Economics RP 181, Williams College, 23. Besant-Jones J. 1999. The Impact of the financial crises Williamstown, Mass., USA on the power sector of transition countries. In Energy after 9. Haggard S. and Webb S. B. 1993. What do we know the Financial Crisis, Energy and Development Report 1999, about the political economy of economic policy reform? 24-32. The World Bank/ESMAP, Washington D.C. World Bank Research Observer 8: 143-68 24. World Energy Council. 1998. The Benefits and 10. Campos J. E. and Salehi Esfahani H. 1996. Why and Deficiencies of Energy Sector Liberalization Volume 2. World when do governments initiate public penterprise reform? The Energy Council. 248 pp. World Bank Economic Review 10: 451-85 25. Energy Sector Management Assistance Programme. 1999. 11. Cook P. and Kirkpatrick C, eds. 1997. Privatization and Global Energy Sector Reform in Developing Countries: A public enterprise reform in developing countries: The World Scorecard. The World Bank, Washington D.C. 73 pp. Bank's Bureaucrats in Business Report. Journal of 26. Müller-Jentsch D. 2000. The Euro-Mediterranean Energy International Development 6: 843-97 Partnership: Sector Policies for the 21st Century. The World 12. Haugland T., Ingeberg K. and Roland K. 1997. Price Bank, Washington D.C. 60 pp. reforms in the power sector: the World Bank's role. Energy 27. Kennedy D. 1999. Competition in the Power Sectors of Policy. 25: 1041-9 Transition Economies. Working Paper 41. EBRD, London. 13. Hawdon D. 1998. Improving the performance of 24 pp. electricity industries in developing countries: Is World Bank policy on deregulation the way forward? In Deregulation of 28. Center for the Advancement of Energy Markets. 2000. 43. Gilbert R.J. and Kahn E.P., eds. 1996. International Retail Energy Deregulation Index 2000: A Scorecard for Comparisons of Electricity Regulation. Cambridge University Measuring State Progress on Electric Restructuring. Press, Cambridge. 500 pp. http://www.caem.org 44. Gray D.R. and Schuster J. 1998. The East Asian Financial 29. Anderson B. and Bergman L. 1995. Market structure and Crisis - Fallout for Private Power Projects. Public Policy for the price of electricity: An ex-ante analysis of the deregulated the Private Sector Note 146. The World Bank. Swedish electricity market. Energy Journal 16: 97-109 Washington D.C. 8 pp. 30. Arocena P., Kühn K. and Regibeau P. 1999. Regulatory 45. Roxas F.Y. The Importance and the Changing Role of the reform in the Spanish electricity industry: a missed Independent Power Producers (IPPs) in the Proposed opportunity for competition. Energy Policy 27: 387-99 Competitive Power Market in the Philippines. Paper presented 31. Green R.J. and Newbery D.M. 1992. Competition in the at the APEC Clean Energy Forum, Bangkok, October 29 to British electricity spot market. Journal of Political Economy November 3, 2000. 12 pp. 100: 929-53 46. Financial Times. 2000 Unravelling Polish PPAs. In Power in 32. Fehr N.H.M. and Harbord D. 1993. Spot market East Europe 39, 11 August: 23-4 competition in the UK electricity industry. Economic Journal 47. Financial Times. 2000. Agreement on spot market not 103: 531-46 enough to get Dominicans out of the dark. In Power in Latin 33. Green R. 1998. England and Wales ­ a Competitive America 62, August: 12-3 Electricity Market? Rep. PWP-060, University of California 48. Financial Times. 1999. Hubco at an Impasse. In Power in Energy Institute, Berkeley CA., USA. Asia 285, 6 September: 6-7 18 34. Energy Sector Management Assistance Programme. 2000. 49. Power Engineering International. 1999. Fast track Privatization, Competition and Regulation in the British Bangladesh Barges. Power Engineering International. Electricity Industry, with Implications for Developing July/August: 71-5 Countries. The World Bank, Washington, D.C. 51 pp. 50. Besant-Jones J. 1996. The England and Wales Electricity 35. Littlechield S. 1994. Competition, monopoly and Model - Option or Warning for Developing Countries? regulation in the electricity industry. In From Regulation to Viewpoint 84. The World Bank, Washington D.C. 4 pp. Competition: New Frontiers in Electricity Markets, ed. 51. Besant-Jones J. and Tenenbaum B. 2001. The California M Einhorn, 125-49. Kluwer Academic Publishers, Experience with Power Sector Reform: Lessons for Developing Boston. 286pp. Countries. Energy Sector Management Assistance 36. Hennemeyer P. 1999. Energy reform and privatization: Programme, Washington D.C. 42 pp distilling the signal from the noise. In Can Privatization 52. Bacon R. 1995. Restructuring the Power Sector - The Case Deliver? Infrastructure for Latin America, ed. F Basañes, of Small Systems. Viewpoint 61. The World Bank, E Uribe, R Willig, 255-307. Inter-American Development Washington D.C. 4 pp. Bank, Washington D.C. 339pp. 53. Global Private Power. 2000. Tractabel set to pull out of 37. Moscote R., Maia S.B. and Vietti JL. 1995. The Power Kazakhstan. Global Private Power, May 11:60 Sector in LAC: Current Status and Evolving Issues. Latin 54. Financial Times. 2000. Brazil: power investors shakeout America and The Caribbean Technical Department Regional begins. In Power in Latin America 56, 3 February: 1 Studies Program Report No. 35.: The World Bank, 55. International Private Power Quarterly. 2000. Third Quarter Washington D.C. Report, McGraw-Hill, New York 38. Chiwaya A., Dube I., Karekezi S., Marundu E., Mugyenzi J. and Mutiso D., 1999. Reforming the Power Sector in Africa, ed. M.R. Bhagavan. AFREPEN, Nairobi. 380 pp. 39. Bastos C.M. and Abdala M.A. 1996. Reform of the Electric Power Sector in Argentina. Buenos Aires. 315pp. 40. Czamanski D. 1999. Privatization and Restructuring of Electricity Provision.: Praeger. London. 152 pp. 41. Lauriol T., ed. 1998. Electricity Liberalisation in the European Union: Law and Commentary. Sweet and Maxwell. London. 196pp. 42. Suding P.H. 1996. Opening up and transition, success and problems. financing and reforms of the electric power sector in Latin America and the Caribbean. Energy Policy 24: 437-45 THE WORLD BANK GROUP The Energy and Mining Sector Board The World Bank 1818 H Street N.W. Washington, D.C. 20433 USA