At a Glance The World Bank FPD Note No. 15 July 1994 Options for Private Power A framework for decisionmaking Denni-s Anderson' m 77Te s52ft to private power means coun- tries have to choose from among sever- al ins-t utional models, ranging from impetus for change the prf.ately owned, vertically integrat- ed uti;tity to theffully "unbuindled" or There are several reasons why countries are seeking to move from state to private 'pure model in which power stations, ownershio: transviission grids, and regional distri- - butior. networks are all privately and - Thefinzancial requtirements of meeting * separately owned. There is often in- new demnands. Demand is doubling tense debate over which model to select every - to 10 years in most developing and wh1at criterion to use. There are regions. Prices averaged less than half _ g some ,-oncerns that choices are too the supply costs in developing coun- some ''oncerns that cho^ces are tootries in 1991, owing to the political .Fw ^ < 4 - ,X. n frequently driven by market ideology regulation of prices. Revenue shortfalls rather than objective analysis of the amounted to over S100 billion per alternatives. To avoid or address these year, and coincidentally were about situat;ons, standard indicators of eco- the same as is required for new invest- nomi perfornance, based on cost- ments. Privatization is seen as a means and price-efficiency and quality of of achwevng price-efficpency where servicc. willprovide some reliable and * Cost atd manragerial efficiency. Large nonideological answers about what electric:il losses. brown-outs, black- best s-,its a particular country-and outs, and poor plant availability are will a.-.so indicate when and where common experiences with state-owned refor7m. is merited at all. 7his Vote out- utilities. and have similarly caused lines i;e reform. options, provides the manaygers and pvate capital to iat- standard indicators which may give prove the efficiency and reliability of some -ciddance on choice, and briefly supphl. review s some of the risks policymakers . M1acroeconomics. The drain on public revenues caused by subsidizing a face i ! any tra nsition period. state-owned utility has also led many countries to look for an alternative Four Models model. There are, broadly, four institutional mode's for a country to choose from. All co: trast distinctly with the most Industry and Energy Department 11- - V-ice P-esidency for Finance and Private Sector Development commion existing approach based on state-owned prices to reflect costs, and this would simultaneously and -controlled enterprises with no private share (i) greatly improve internal caih generation in most ownership-the point of departure for the majority countries and (ii) lay the basis for private capital to of developing countries. The alternatives are: be raised in stock and bond markets, and through direct investment. .llodel 1. The traditional model of the vertically inte- grated state-owned public utility, operated as a qua- Mlodel I would thus lead to financial self-sufficiency si-independent public corporation on commercial in the industry and eliminate the budgetary distress- principles. Regulation is independent of-arm's es that the poor pricing policies of the public utili- length from-government, the utilities, and special ties have caused in many countries. But it remains, interests. This is the model followed by rnost Euro- nevertheless, a "regulated monopoly," and lacks the pean utilities, including the U.K. until 1989. Korea, element of competition and the incentive for mana- and Thailand. It is not, of course, strictly a private gerial and cost-efficiency in the provision of sup- model, though what has differentiated it from the plies. Second, regulated moncoolies are often wed- models followed by most developing countries is ded to existing approaches and inhibit investment in independent regulation and that it has attempted to newly emerging electricity supply technologies. "simulate" the results of market competition, e.g., through the adoption of commercial pricing policies Model 1I overcomes these problems by admitting (ideally marginal cost pricing). independent generation on the supply system. The model introduces competition in new generation, Model II. Like MIodel I, but with financial resources and attracts the additional financial and managerial being raised through private share ownership, resources associated with direct investment. In addi- bonds, and private borrowing, and with competition tion. it is associated with greater accountability be- being introduced by private power producers (the cause of the requirements of independent audits and U.S.). the public scrutiny that comes with share owner- ship. The increased accountability extends not only Model III. Like Model I, with competitive procure- to cost-efficiency in operations and investment, but ment of new generation plus a contract or com- also to public issues such as compliance with taxa- mon carriage for transmission. Countries in the tion and environmental laws and the provision of process of moving to complete vertical separation universal service. may pass through this phase (the U.S. since 1992, Mexico). MVodels III and IVare institutionally complex, and will not always be the most aopropriate starting Model IV Complete vertical separation of genera- point for many countries. Moziel III opens plant dis- tion, transmission, and distribution, with fLill priva- patching schedules to compet-'ion and this is clearly tization of each (the U.K., Argentina, Chile. Peru). a force for cost-efficiency. In addition, it should lead to improved tariff structures: prices would, in theory, Benefits of change be bid down to marginal fuel rnd operating costs at The choice of one model should not preclude evo- off-peak, and up to marginal c.pital plus fuel and lution into another. A country can begin to reform operating costs at peak. Seasonial pricing would be its utility using Model I or II, and evolve to Model III encouraged on predominantly hydro svstems, and or IV None of the models can work without a rea- contracts between supplier and consumer woLld sonable degree of price-efficiency, or thus without lead to prices correctly reflecting costs according to solving the financial problem, and for this reason voltage level. If prices did nor reflect such cost lev- alone all have their merits. Moving from the state- els and structures, this would he an unmistakable owned enterprise to Model I, for instance. requires sign that competition was lackling. 2 Model IV has two advantages over Mfodel III. First, it table indicates where possible the standards that are allows for the franchising of dlistribution services. typically achievecl in best practice situations. Hence, companies that believe they could provide these services more efficiently are able to contest Managing risks in transition current arrangements. Second. distributors will be Table 1 concentrates on static' measures at a point better placed than others in the industry to judge the in time only. But for most countries the path for commercial merits of decentralized electricity pro- duction, such as cogeneration or smaller-scale gen- eration owned by the distribution companv itself or Table 1. Measures of Efficiency* by private investors. NModularity in fossil-fired plants and in renewable energy supplies is opening up Cost-efficiency manv possibilities in this direction. 1. Reserve plant margins (20%) 2. Plant availability factors (85 to 90%) The biggest economic and financial gains might well 3. Total systems losses (6 to 10%) be realized by going from the state-owned enter- * coal (2.300-2-800) prise to MVIodel II. Some countries might prefer this * oil (2,300-2.700) course if only because they could use the public * gas turbines (2,500-2,000) utility to take the lead in extending regional and * combined cvcle (1,700-2,000) rural services. Another option might be to combine iVodel II with elements of Model IV (such as fran- Priceelfficilency chising) in order to encourage managerial and price- 2. Self-financing ratio (>25%) efficiency in supply and investment in decentralized 3. Billing efficiency: thefts (<2%), collection period (<30 to generation. 40 days) 4. Percent share, equity, and bond finance Objective basis for choice 5. Demand management: -_bictiv b* load factor improvement policies The success or failure of any institutional model can * power factor improvement be objectively determined by measuring its economic * peak load pricing efficiency. The advantage of the efficiency criterion v pricing by voltage level is that it immediatelv suggests a range of indicators by wvhich the success of a policy can be meeasuredl. Provision of ser-vice p Y ~~~~~~~~1. Loss of loaid probabilitN, i- hours per year) Some indicators are shown in Table 1. They are 2. Percent of public served (rises with per capita income neutral among institutional models. ancl are not in- level and costs and benefits of electrification: long-term fluenced by ideological factors. If an electricity in- gooal is universal provision) dustrn scores well on all of them, there might be no 3. compliance wvith environmental policies (varies and reason to reform it. no matter wh1ich modlel it con- evolves vith policics. hut high levels of abatement can .if it is a pulic enterprise. he achieved through control technologies for PM. NOx. forms to-even If It Is a publlc entet-prise. wand S02 or use of- clean fuels Sucl as gas) The indicators have been divided into three (groups to reflect different aspects of economic efficiency: manCe dr.d)d1`1 'tCC hebv . WorldI 13ank enurgv staff. Februare whether the investment and operating costs are be- ') 4 ing, minimized bh the new moldel (cost-efficiencv) -TeSC rItes Ot lrtLirn arc loe titan tlhe r-tLlirns of 20 percent otten q(ioted by private coinpanies. \vhich incluCle allowoances whethler the level and structure of prices reflect the io- cost cscalation andci slippagc>e and aklo for Countrv risk in level and structure of costs (price-efficiency); and the case of foreign investors. The expected or average real whether service is being provided in wavs satisfacto- financial returns are very likely muech lo\ver. comparable to the ry to the public-a criterion which includes environ- average real financial return of 8 to 12 percent. mental factors as well as the provision of kWh. The I 3 moving from one model to another may be as im- * rising demancds for new capacity. The amount of portant as the choice of the mociel itself. ancd will new investment required in the clecacle after varv with the situation. Some Countries might initial- privatization will probably exceecl the size of the lV prefer Model II (the traditional utility operating on existing system. New investors may he more at- commercial principles. with competitive procure- tracted to supplying new capacity than to acquir- ment of new generation), and shift to Models III or ing and rehabilitating the existing system. If so, it IVat a later time. Others might be concemed with would be at least a decade before the majority of rigidities setting in once a decision has been made the system could be privately owned, even in to set up a competitive arrangement between the theory. public utility and private producers under Model II. * the large numbers of people and small commer- and therefore might prefer to go clirectly to Model III cial consumers without service. Roughly 2 billion or IVI people do not have access to electricity in devel- oping countries, and with population growth and The path chosen will depend on the capacity and urbanization the demand for new connections will willingness of the capital markets to support the be high. reforms. In developing countries, the stock markets * the need to raise prices by 50 to 100 percent or presently have a capitalized value of about S800 more in many countries. billion and are raising an additional S65 billion each * fears of nationalist backlashes against foreign year-only about half the electricity industry's annu- ownership. Local capital markets will not be able al new investment requirements of $120 billion, and to absorb the industry fully, and finance will be of course there are the financial demands of other forthcoming from foreign buyers only if political sectors to be met. Even when aggressive policies are and economic risks are low. put in place to change the legal framework and cor- poratize the industry, the transition to full private The danger is that if these problems are not ad- ownership cannot proceed faster than the private dressed privatization may not be sustainable, and sector's ability to finance it. there may be calls for renationalization-a possibili- ty that the investors will doubtless have in mind. As In turn, availability of finance will be affected by the one private producer has put it, their fear is that following: 'nationalization will not be an intermediate step between privatization and privatization. but that * the inheritance, in most countries, of capacity privatization will be an intermediate step between shortages, overloaded distribution systems. large nationalization and nationalization. All the more losses. thefts of supplies often amounting to over reason, therefore, to choose the model and the path 20 percent of electricity prodluced, much generat- of reform carefullv. ing capacity in disrepair. and numerous manageri- al problems-including overmanning and lealing Thi.s \otc has cspeciallv henclitcd tlrnT maturiais and curmments pro- with redlundlancies-to be sorted o>ut. This w,ill idi.d ih) 1Ocrn;:rd1 rcnncnhiirn. lohn iB.lni-Kones. and Peoter Cordutikes. leter private acquisition of the existing assets un- less their prices are bid clown to a low level. Dennis Anderson Senior Advisor. tndeustrLr and Energy Dcpartrnnt. 'I'his series is published to share ideas and invite discussion. It covers financial and private sector development as well as industry and energy. The views expressecd are those of the author(s) ancd are not intendecl to represent an official statement of Bank policy or strategy. Comments are welcome. Please call the FPD Note line to leave a message: 202-458-1111: or contact Suzanne Smith, editor, Room G8105. The World Bank. 1818 H Street, NW, Washington, DC 20433, or Internet address ssmith@worldbank.org. 4