Report No. 1 5063-AR

ArgentinaReforming Provincial Utilities:Issues, Challenges and Best PracticeJune 6, 1996

Infrastructure DivisionCountry Department ILatin America and the Caribbean Region

4 ' '2' ' ,,.''0 , :, -:,,,:f... .

, ,.,_ .

* ou~v r - h W -, - ,, ,.,*

CURRENCY EQUIVALENTS

Currency Unit - Argentine Peso (A$)

US$ = A$1

WEIGHTS AND MEASURES

The Metric System is used throughout this report.

FISCAL YEAR

January I to December 3 1

ACRONYMS

AyE Agua y EnergiaCAMMIESA Compafiia Administradora del Mercado Mayorista Electrica Sociedad An6nimaCFE Federal Council of Electric EnergyCOFAPYS Consejo Federal de Agua Potable y SaneamientoENOHSA Ente Nacional de Obras Hidricas de SaneamientoES The National Energy RegulatorENRE Ente Nacional Regulador de la ElectricidadETOSS Ente Tripartito de Obras y Servicios SanitariosGUMA Gran Usuario MayorGUME Gran Usuario MenorLRMC Long-run marginal costNARUC National Association for Regulatory Utility CommissionersNRRI National Regulatory Research InstitutePPAD Precio de la potencia puesta a disposici6nRMBA Reforma Metropolitana de Buenos AiresSADI Sistema Argentino de Interconexi6nSRNAH The Secretariat of Natural Resources and Human EnvironmentSPAR Provincial Service for Drinkable Water and Rural SanitationSWR Subsecretariat of Water ResourcesPWS Public Works SecretariatW&S Water and Sanitation Services

PREFACE

This report has been prepared by a team led by Antonio Estache (LA1IU) based on the findings ofmissions to Argentina in February 1995 and in June 1995. The core team included Claude Crampes(Universitd de Toulouse, Institut d'Economie), Marianne Fay (AF lET), Walter Garcia-Fontes (UniversitatPompeu Fabra, Barcelona), Frannie Humplick (PRDEI), and Thomas-Oliver Nasser (MIT and Institutd'Economie Industrielle, Toulouse). The report also benefited from background papers by XavierFreixas (Universitat Pompeu Fabra, Barcelona), Felix Helou (Consultant), Martin Rodriguez-Pardina(Consultant), Ben Shin (PSD) and Warrick Smith (PSD). Also we would like to thank Secretary Zapata,Mr. Segnana and Mr. Vega, our main counterparts during the mission, for providing us with atremendous amount of information and allowing us to meet key counterparts from the Provinces ofBuenos Aires, C6rdoba, Mendoza, Neuqu�n, Santa Fe, Santiago del Estero, Tierra de Fuego andTucum�n. Finally, we would like to thank Secretary Bastos and Undersecretary Rotaeche for very usefuldiscussions and insights.

Background Papers Supporting the Report

Fay, M. (1994), "Infrastructure and Growth in Argentina", mimeo, the World Bank, LAIIN

Fay, M. and A. Estache (1995), "Regional Growth in Argentina: Determinants and Policy Options",mimeo, The World Bank, LAI IN

Freixas, X. and W. Garcia-Fontes (1995), "Infrastructure Financing for Argentina's Provinces: Issuesand Options", mimeo, the World Bank, LA1fl�4

Humplick, F. (1995), "Infrastructure Performance in the Provinces of Argentina", mimeo, The WorldBank, LAlIN

Humplick F., and T. 0. Nasser (1995), "Risk in Provincial Infrastructure Provision: An Investor'sPerspective", mimeo, The World Bank, LAlIN

Rodriguez-Pardina, M. and F. Helou (1995), "Comparaci6n de Marcos Regulatorios en Argentina",mimeo, The World Bank, LAIIN

Smith, W. and B. Shin (1995), "Regulating Infrastructure: Funding Regulatory Agencies", mimeo, TheWorld Bank, LAlIN

Smith W. and B. Shin (1995), "Regulating Infrastructure: Perspectives on Decentralization", mimeo, TheWorld Bank, LAI IN

ARGENTINA

REFORMING PROVINCIAL UTILITIES:

ISSUES, CHALLENGES AND BEST PRACTICE

TABLE OF CONTENTS'

Executive Sum m ary .............................................. i

Chapter 1: INTRODUCTION .......................................... ,.1The Need for Credible Provincial Commitment .................... ....................... IWhat Competition in Utilities Entails ........................................... , . 2What Contract-Based Regulation Entails ........................................... 5Does the Provincial Choice of Regulatory RegimesMatter to a Potential Investor? ........................................... 6What Decentralized Regulation Entails ........................................... 7

Chapter 2: REFORMING ELECTRICITY DISTRIBUTION IN THE PROVINCES ................ 9Overview of the sector's Organization ............................................................... 9The "Regulators" ....................... 14Contracts as the Main Regulatory Tool ........................... 17How Competition Works ........................... 18Pricing ...... , .. ,,,,,.,.... 18Investment and Tariffs ................................. 26Summary of the Main Recommendations ................................. 29Generation ................................. 29Transmission ................................. 29Distribution ................................. 29Regulation ................................. 30

Chapter 3: DELIVERING WATER AND SANITATIONSERVICES IN THE PROVINCES ......................................... 31Overview of the Sector's Organization ......................................... 31The "Regulators" ......................................... 32Contracts as the Main Regulatory Tool ......................................... 33How Competition Works ......................................... 36

lThis report was produced under the supervision of Mr. Gobind Nankani, Director; Mr. Asif Faiz, Division Chief,Infrastructure and Urban Development; and Mr. Danny Leipziger, Lead Economist, Country Department I, Latin Americaand the Caribbean Regional Office. The peer reviewers for the early drafts were Messrs. 1. Kessides, PSD and L. Guasch,LATAD.

Pricing .............................................. 37Investment and Tariffs ............................................. 40Summing up the Main Recommendations ................ ............................ 41On Tariff Design ............................................ 42On Contract Selection ............................................ 42On Institutional Reform ............................................ 42

Chapter 4: DEVELOPING A PROVINCIAL REGULATORY CAPACITY ....... 43Breaking with the Past ...................................................... 43Constraints and Trade-Offs ...................................................... 44Sectoral Breadth of Authority ...................................................... 45Is Representation Needed in Each Municipality? ................................................... .. 46Minimization of Regulatory Demands ....................................................... 46Size of Agency Staff ...................................................... 47Strengthening Regulatory Capacity ...................................................... 48Funding the Regulatory Agency ...................................................... 49Organizing Interprovincial Cooperation on Regulatory Matters ............................... 50Summing Up ....................................................... 50

TABLES

Table 1.1 a Status of Provincial Reforms in Electricity ....................................................... 1Table 1. lb Status of Provincial Reforms in W&S ...................... .. 1..............................ITable 1.2 Summary of Main Concession Contracts ....................................................... 4Table 2.1 Structure, Financing and Functions of ENRE, the National

Entity for the Regulation of Electricity ........................ .............................. 15Table 3.1 Impact of the W&S Reform in the Greater Buenos Aires .......................... ................ 31Table 3.2 Main Provincial Water Contract ...................................................... 34

BOXES

Box 1.1 How to Check if the Reform is being Captured by the Bidders? ............. ...................... 3Box 1.2 Comparing Concession Process in W&S and in Electricity ......................................... ... 5Box 2.1 Who are the Main Generators? ...................................................... 12Box 2.2 Who Transmits where in Argentina? ...................................................... 13Box 2.3 What did the Concessionaires do with their Property Rights

in Distribution? ...................................................... 14Box 2.4 How are the Marginal costs Computed? ...................................................... 19Box 2.5 Dealing with the Seasonality of Prices ...................................................... 23Box 2.6 A Comparison of Tariffs across Provinces ...................................................... 24Box 3.1 The Timing of the Privatization Process in Greater Buenos Aires ............... ................. 33Box 3.2 Goal-Based vs. Process-Based Contracts ...................................................... 35Box 3.3 What Contract Renegotiation Reveals ...................................................... 36Box 3.4 What Yardstick Competition Entails ...................................................... 37

Box 3.5 Why does the Concessionaire Want Meters Everywhere ............................................. 39Box 3.6 What's Wrong with the Average Price of Water in

Argentina's Provinces? ................................................ 40Box 3.7 Willingness to Pay for Sewerage ................................................ 41Box 3.8 Dealing with Social Concerns ................................................ 41Box 4.1 How to Appoint Regulators? ................................................ 44Box 4.2 Consumer Representation ................................................ 48

FIGURES

Figure 2.1 Financial Flows ................................................ 11Figure 2.2 Marginal Cost and Inverse Demand ................................................ 20Figure 3.1 ................................................ 39

ARGENTINA - REFORMING PROVINCIAL UTILITIES: ISSUES, CHALLENGES ANDBEST PRACTICE

EXECUTIVE SUMMARY

This report analyzes the reform needs and options in electricity distribution, and water andsanitation services (W&S) in Argentina's provinces. The main focus is on the regulation of privateoperators since many provincial governments are considering concessioning all or part of theseservices to the private sector to improve quality and rely more on private financing of the sectors'expansion needs. Tariff design and institutional concerns are emphasized in a detailed analysis of thespecific regulatory framework of both sectors, including some issues raised by federal regulationrelevant to the decisions of potential investors in provincial utilities.

Main conclusion. The report strongly endorses the general concessioning strategy followedby most provinces. This strategy has already generated substantial gains in utilities under Federalcontrol. In electricity for instance, tariffs have slightly decreased on average since privatization andquality of service have improved significantly. For instance, the average number of annual serviceinterruptions was cut from at least 7.6 to 5, and the average duration from 23 hours/year to9.9 hours/year. However, not all provinces will be able to follow a strategy relying on privateinvestors as they vary tremendously in terms of potential rate of return and commercial as well as non-commercial risks (including political and fiscal risks). Some of the poorest provinces combine lowpotential return and high risks' levels and are unlikely to be very attractive under deals and regulatoryarrangements that try to shift most of the risk onto the private investors. For these provinces or forthose not wishing to transfer outright the W&S and electricity distribution services to the privatesector, the report offers alternatives. It suggests, in the short run, the commercialization of theseservices through management and service contracts with private companies can serve to establish thecredibility of the provincial commitment to reform. But this should be viewed as a first step towardsthe implementation of concession as it gives investors a chance to obtain more independentinformation on the value of the assets for which they would be bidding

The keys to successful provincial utilities reforms. The National Reforms demonstrate thatto maximize the gains from the "privatization" strategy, the following steps are needed:

* make the most of what competitionfor the market allows,

* give an incentive to new owners to remain concerned with the public interest;

* assign rights and obligations in the contract as clearly as possible,

* anticipate the potential needs for renegotiation;

* define the nature and form of property to maximize accountability,

* do not underestimate the importance of tariff design; and

* prepare and develop the provincial regulatory capacity as carefully as the contract.

ii

Do not underestimate what can be achieved through competition for the market.Whether in the area(s) of operations, management or specific services, competition for the service iscrucial to the success of the current reform efforts. Competition and increased private sectorinvolvement also reduce arbitrary political interference in price setting or employment decisions, forinstance. Although the specific form of private sector involvement in each province will have to betailored to fit local needs, constraints and preferences, the introduction of competition, if implementedcorrectly, will cut total service costs, pass a large share of these cuts to all consumers and improveresponsiveness to users. The success of the approach (in terms of fiscal impact but also in terms ofservice quality and price) depends a lot on the design of the bids to assign the operator's role in aconcession, management or service contract. Some of the provinces are clearly aware of theimportance of this step in the reform process: it ensures the credibility of the provinces'announcements and their commitment to change. For instance, the province of Santa Fecommissioned an independent assessment of the value of the net assets of its utilities (done by aninternationally recognized external expert). This provided independent information on the potentialfiscal impact of the reform. It also revealed up front as much reliable information as possible topotential investors. This type of assessment can also avoid some uncertainties in terms ofrehabilitation needs and reduce the risk of unexpected tariff adjustments as those approved for AguasArgentinas over and above what was agreed in the original contract.

Ensure the managers' and workers' accountability. If property is distributed, the way it isdone can also contribute to the success of the reform. Shares can be sold on the stock market towiden the dissemination of property and hence of accountability as was once considered (buteventually rejected) for Aguas Argentinas. Shares can also be allocated to workers and employees, asfor electricity transmission under federal jurisdiction, giving them an effective incentive to support thereform and to act in the interest of the enterprise. But there are risks in spreading accountability toothinly. This is why it is generally suggested to allow the bidding of packages of shares large enough toallow control by one major interest i.e., to have a strategic investor.

Clearly spell out the rights and obligations of all parties. Under the strategy adopted bythe provinces, contracts (concession, management or services) are the main regulatory instrumentduring the tenure of the private operator. Where and when governments do not have a long trackrecord in dealing with the private sector, contracts need to be drawn as tightly as possible to reduceopportunities for discretionary government actions. Contracts are intended to be enforced accordingto their terms. The provincial reforms will only work if the provincial governments recognize thatonce a private operator takes over, this operator is in charge within the specific terms of its contract.

Anticipate the potential needs for renegotiation. Even if a province has well prepared itscontracts, unforeseen events will happen. This means that even if contracts should be prepared asdocuments that are not intended to be modified and include adjustment mechanisms to avoid the costsand uncertainty of renegotiation, under very specific and very limited circumstances, a limited degreeof contract flexibility may be good public policy. To ensure that the contract is a credible regulatoryinstrument, contract modifications should be based on some fair and workable rules based on clearlyspelled out policy criteria (discussed in the report). These rules should be agreed upon by all involvedparties, and the guarantee of this agreement should be provided by an independent regulator as anhonest broker. Unilateral modification of the rights defined in the contract or the bidding documents(say through request to accelerate investment programs or impediments to tariff adjustments) without

iii

full compensation is tantamount to expropriation and defeats the purpose of the reform. It is the riskof arbitrary government behavior of this kind that historically created the biggest risk for privateinvestors in infrastructure and that continues to deter many deals.

Keep the new private monopolies working in the public interest. Once the contracts havebeen awarded, it may be difficult for the government to get the private operator or managers to focuson the interest of consumers as much as on their own interest. To maximize the incentive for efficientbehavior during the concession, the contracts can require that new auctions be organized at regularintervals as it was done in the case of the electricity distribution concessions orchestrated by theNational Govemment. Another incentive for provincial monopolies to maintain their concern for theinterest of consumers is to rely on some formal comparison of performance across provinces. Thesecomparisons should be widely disseminated (in the media) to increase the public pressure foraccountability. These comparisons can also be used more formally by regulators to push prices totheir lowest possible level ("yardstick competition"). This requires a strong inter-provincialcoordination to standardize information which could be organized by the National Regulators.

Get the tariffs design right. All reformers recognize that the design of tariff formulas are atthe core of effective regulation. However, few appreciate its complexity and relevance to the long-term sustainability of the provincial utility reforms. The report discusses how tariff design matters toensure the best allocation of scarce provincial resources. In the context of this analysis, it identifies afew issues that deserve their immediate attention. Some of these issues are under their direct controland should be handled as part of the reform process. Some are under the responsibility of the NationalGovernment but need to be monitored by provincial reformers to allow them to anticipate theconcerns of potential investors in provincial utilities.

In Electricity, the provinces trying to introduce more efficient pricing of distribution services(including an incentive to maximize productivity gains) should adhere to the tariff methodologycontained in the national electricity law with some adjustments. Chapter 2 shows that:

* If regulation continues to be based on price capping (aims at giving incentive to minimize costs),there is a need to very quickly define a methodology and gather the information needed tocalculate the productivity gains that should be passed on to consumers.

* There is also a need to review the current tariff methodology in order to make sure that the interestof "captive consumers" are protected. The current approach creates a distortion due to thepossibility of direct contracts between "large users" and "generators". The consequence of thesecontracts is that captive consumers (i.e., those who are not large enough to be able to buy withouthaving to go through the distribution companies) may have to pay much higher rates because theyare paying for more expensive wholesale contracts transferred at the time of privatization andlarge users opt out to negotiate cheaper contracts in the wholesale market. This could lead todifficult political problems as "privatized" tariffs may end up increasing significantly in relativeterms for a large share of the consumers. This could be avoided by allowing some more pricingflexibility and some degree of discrimination in prices.

In addition, the provincial governments will have to monitor the developments in transmissionpricing, one of the very few serious outstanding issues in the otherwise very impressive NationalSector Reform. Better and clearer rules (based on economic benefits rather than on energy flows) are

iv

needed to avoid the type of difficulties recently met in discussions of the construction of a fourthtransmission line in the Comahue corridor. Existing rules have a bias towards underinvestmentbecause they fail to assign a clear responsibility for the payment of the construction costs of anyexpansion. Since transmission is the physical instrument to guarantee competition in generation andsupply, this underinvestment is increasingly a concern to potential investors in distribution. Moreover,as more and more provinces privatize their distribution companies, the relationship between thetransmission concessionaires and the distribution concessionaires is likely to change, and the tolls mayhave to be revised to ensure consistency with economically efficient pricing rules and revenueadequacy.

In W&S, Chapter 3 shows that the regulatory challenge is much wider and complex:

* A radical tariff reform is needed to achieve efficiency in the use of water and to finance growinglong-term investment needs, particularly in the sewerage systems. It will involve increasedmetering, eliminating the current tariff for unmetered consumption and redesigning the fixed partof the current two-part tariff used for metered consumption, since it currently leads to multipletypes of cross subsidies, and since it fails to provide much incentive to minimize cost and to investin the expansion of the sewerage system.

* This tariff reform will have to be implemented as part of a wider reform which should aim at theadoption of commercial practices in the sector. This means that the provincial governments willhave to give up many of the controls they are now imposing on their public enterprises and aim atregulatory simplicity when identifying what needs to be controlled/regulated and what need not be(and this will require strong coordination with the National Government).

* The concerns with the high level of uncertainty regarding the asset value and the costs of therehabilitation needs expressed by potential private investors interviewed by the Bank are such thatsome of the provinces will have to consider a slower pace of reform, and will have to focus onshort-term management contracts with an option to transform them into concession contracts inthe longer run. It may also be worth considering the bundling of W&S assets across provinces, ortheir bundling with other assets such as electricity to reduce the overall risk faced or perceived byprivate investors. None of this should affect the efforts to reform tariffs and increase metering.

Develop the provincial regulatory capacity. The most difficult provincial challengeaddressed in the report may be the need to monitor that the behavior of the privatized monopolies areconsistent with the terms of the contract as well as with public aims. This monitoring is neededbecause concessionaires will have strong incentives to behave inefficiently when left unsupervised.The creation of regulatory bodies or the use of the strength of law and courts will be an importantelement in the success of the provincial reform of utilities and Chapter 4 focuses on this exclusively.But there is a need to distinguish between technical and economic regulation in both sectors.Economic regulation does not have the same goals as technical regulation so that its organization isnot necessarily identical to the technical one. For the electricity sector, the difference can beillustrated as follows: at the terminal nodes of the interconnected electric network there is little needfor technical control, but it is a place where provincial distributors can try to exploit their exclusiveposition in front of captive consumers and hence economic regulation may be needed even if technicalregulation is not. Therefore, the skills needed to develop the provincial regulatory capacity are notsimply a recycling of the skills needed to provide the service in a public enterprise. Some of the staff

v

of the former public utility will have to be dismissed and new staff will have to be recruited tointroduce these new skills in the provincial regulatory entities.

Aim at a single independent sector specific agency. While the Report shows that someeconomic regulatory responsibilities may be justified at the provincial level in electricity and W&Seven when technical regulation is not needed, it also shows that if each province must have its ownregulatory capacity, the local entity should be kept to a minimum and only have specific prerogatives,essentially auditing and reporting to the national regulation body. This is particularly important inelectricity because technical dependencies within the network are very strong. Moreover, therecognition of a role for provincial regulation is not an endorsement of the creation of multiple sectorspecific regulatory agencies. The international experience suggests that most provinces would bebetter off with a single independent entity for all utilities. Chapter 4 explains in detail how to achievethat independence in terms of nomination of regulators and financing of the agency.

Do not simply staff the agency with the employees fired by the private operator. Thestaffing level of the agency should be modest and its composition should depend on the requirementsof specific tasks assigned to the regulatory agency. As discussed in the Report, there are many goodpossible criteria for selection (i.e., technical excellence, political representation, ...). With a fewexceptions, the experience considered relevant in the staffing process at the national level was basedon engineering aspects of the sector while economic regulation, the main purpose of these entities, isquite different from technical regulation. This suggests that the selection criteria for the regulatorsshould include expertise in economic fields. While the skills needed to be a good regulator are hard tofind, they are very similar in both, the water and electricity sector, and hence the same resources canbe shared within a single provincial regulatory entity. This means lower resource needs (not only interms of staff but also in terms of equipment) and hence a lower burden on the taxpayers or thebeneficiaries of the services. There are other benefits. For instance, it facilitates learning betweensectors, ensures consistent approaches to central policy questions, and makes the regulator lessvulnerable to industry or political capture. This lower vulnerability enhances the credibility of theagency and reduces risks for investors.

Spell out the implementation strategy. The strategy to implement this regulatory agencycan be flexible. If no agency has been established yet, a core multisectoral framework can be set upfirst and then sectors can be added to the core entity when appropriate. There is no need to wait untilany contract has been signed. The creation of an entity can be instrumental in assisting in theimplementation of the concessioning process. If one agency has already been established, it should notbe too difficult to expand the jurisdiction of the existing agency. This depends of course on howsector specific in design and composition the initial agency is. Finally, if several agencies have alreadybeen created, the best option is to develop a strategy for merging the agencies after an initial period.This strategy tends to be the most difficult and will often be resisted by both the existing regulators(concerned about losing their autonomy and possibly their job) and investors (often preferring thecounterpart they know).

Next Steps. The main conclusion of this report is that there is no single blueprint for theprovinces but that there is strong evidence of a need to further understand regulatory objectives andestablish regulatory rules and regulatory behaviors whatever the specific strategy adopted by aprovince. The ultimate aim is better and cheaper service provisions to the Argentine consumer,

vi

whether she/he lives in Bs.As. or Salta or Tierra del Fuego. The Report also aims to be a discussionreport, namely a vehicle to raise issues and begin the public dialogue necessary to extract the benefitsof a private sector participation in key utilities in Argentina's provinces.

Chapter 1: INTRODUCTION

1.1 Objectives of the report. Most provinces Table 1.la Status of Provincial Reforms in Electricityin Argentina are concerned with the high cost and _ y 1996)poor quality of service in electricity distribution Already concessioned Catamarca, Entre Rios,Formosa, La Rioja, Rio Negro,and water and sanitation (W&S) as well as with San Juan, San Luis, S. Delthe lack of public resources for their rehabilitation Estero, Tucumanand expansion. The report analyzes the reforn At the bidding stage Jujuy, Saltaneeds and options in these two sectors in With privatization and Misiones, NeuquenArgentina's provinces. The main focus is on the regulatory frarneworkregulation of private operators since many definedpregulatio gofprivaterr peratos ansincern m With regulatory Bs.As., Corrientes, Misiones,provincial governments are considering framework pending Santa Feconcessioning these services to the private sector No legal framework C6rdoba, El Chaco, Mendoza,(see Table l.la and 1.lb on the current status) yet or irrelevant Santa Cruz, T. d. Fuego (Chubutand are requesting technical assistance in and La Pampa will keep theiridentifying the main regulatory issues they will cooperatives).have to address. The lessons of the nationalexperience with contract-based regulation and Table 1.lb Status of Provincial Reforms in W&Stariff design as well as institutional issues are 1996_______________ s &

particularly detailed throughout the report--much Already concessioned Corrientes, Formosa, Santa Fe,more so than in previous recent Bank publications TucumAnon these two sectors.' This detailed analysis is At the bidding stage C6rdoba,needed to show explicitly the linkages between: With privatization and S. del Estero, Mendoza(i) tariff design on one hand, and investment regulatory frameworktargets and financing on the other hand, and definedWith regulatory Catamarca, Jujuy, La. Rioja,(ii) national and provincial regulatory issues. framework pending Neuquen, Salta, San Juan

No legal framework Bs.As., El Chaco, Entre Rios,1.2 Road Map. Chapter 2 covers the main yet or irrelevant Rio Negro, San Luis, Santaissues in the privatization of provincial electricity Cruz, Tierra del Fuego;distribution companies. Chapter 3 deals with (Chubut, and La Pampa willW&S services. Chapter 4 discusses the probably maintain theirinstitutional demands of the regulation of private cooperatives)operated utilities. The remainder of this firstchapter reviews the main general policy issues the The Need for Credible Provincial Commitm tprovincial governments will have to address and 1.3 If efficiency in the allocation of scarce

that apply to both sectors: provincial resources and better responsiveness to

* what limited credibility entails user needs are the main policy objectives of the* what competition entails provincial governments, the report endorses the* what contract-based regulation entails general concessioning strategy being followed by* what decentralized regulation entails. most provinces. But not all provinces are equally

attractive to private investors as they varytremendously in terms of potential rate of return

See for example, World Bank Report No. 35 The Power and commercial (some have low potential returnSector in LAC: Current Status and Evolving Issues by R. A. and high risks level) as well as non-commercialMoscote, S. B. Maia, and J. L. Vietti, June 1995.

2

risks (including political and fiscal risks). The these rights (say through request to acceleratereform strategy may have to be tailored to the investment programs or impediments to tariffconstraints imposed by these differences. So what adjustments) without full compensation isare the short run alternatives to concessions? tantamount to expropriation. It is the risk of

arbitrary Government behavior of this kind that

1.4 When a province does not succeed in historically created the biggest risk for privateawarding a concession contract (or does not wish investors in infrastructure. Where and whenthe immediate transfer of the service to the private nnvestors do not as thtyhe Goverinlent, contractssector), the report suggests not to give up on need to be drawn as tightly as possible to reducetrying to introduce more incentives to cut costs in opportunities for discretionary actions by thethe sector but to try instead to rely, at least in the Govemmentshort run, on some form of commercialization ofthese services through management contracts: this 1.6 This suggests that the success of thecan serve to establish the credibility of the privatization process depends closely onprovincial commitment to reform in the view of . how property rights are defined andprivate investors before concessions can become a allocated initially (i.e., will the Governentrealistic option. This gives investors a chance to really give up control in exchange forobtain more independent information on the value efficiency and financing?); andof the assets they would be bidding for (a major * under what circumstances could these rights

issue in the W&S sector) and get the be eventually reallocated (i.e., what are thecommercialization of the service going withoutpulcoiycrtiahtcudledoa

dea2 public policy criteria that could lead to adelay. *2renegotiation of a contract ?).3

What Competition in Utilities Entails 1.7 In the national level reforms, the overall

1.5 It seems clear that provincial governments allocation of property rights to maximize the gainshave generally accepted to introduce competition from competition is the outcome of the followingin their local service monopolies through the three prong strategy:organization for competitive bidding. But the . assign property rights as clearly as possible inmain lesson from the national reform experience the contract;that Argentina's provinces may not have yet beenable to intemalize fully is that the key to * define the nature and form of property tosuccessful competition is to assign property rights maximize the manager's accountability; andon all the resources unambiguously. In other . give an incentive to new owners to remainwords, the Government has to give up control efficient throughout their tenure.over the service and recognize that once a privateoperator takes over, the private operator is incharge within the specific terms of its contract 1 8 The first prong was to rely on sealed-bidswith the Government. Contracts are intended to to assign the operator role in a concessionbe enforced according to their terms. This is what contract for a specified period (electricity, water,creates property rights. Unilateral modification of

3 The existence of market failures such as2 To avoid the creation of an information monopoly in environmental concerns often limits the extent to

favor of the winner of the management contract, clear which property rights can be shifted from thedata publications requirements should be spelled out in Government to a private operator.the contract.

3

etc.). As discussed later, the specific design of the 1.11 The third prong is relevant because thebids is a key determinant of the success of reform. provincial services at stake are local naturalSome of the provinces are clearly aware of the monopolies. Once the concessions have beenimportance of this step in the reform process. The awarded, it may be difficult for the Government towater company in Santa Fe, for instance, get the private operator or managers to focus oncommissioned an independent assessment of the the interest of consumers as much as on their ownvalue of its net assets (done by an external expert interest. Once more, the national reform of thein whom prospective investors had confidence) to electricity sector shows how to reduce the risk.reveal up front as much information as possible to To maximize the incentive for efficient behaviorpotential investors. This avoided some (but not during the concession, the concession contractsall) of the uncertainties in terms of rehabilitation awarded by the National Government in electricityneeds observed in the privatization of Obras require that new auctions be organized at regularPfiblicas in Buenos Aires. (Table 1.2 gives a intervals according to a procedure explained inchecklist of the main features that contracts in any paragraph 2.32.4 Basically, while the electricityprovince should cover). distribution concessions have a duration of 95

years, they are divided into management periods1.9 Just as important, all the requirements of of 10-year (except the first one which lasts for 15the bidding documents have to be internally years) which are allocated every 10 years.consistent and reasonable. They should bechecked for inconsistencies between the tariffs andthe investment and rehabilitation requirements. Box 1.1 How to Check if the Reform Process is beingThis may have been an issue in the bids organized Captured by the Bidders?

for the water concession in C6rdoba. Any serious discrepancy between various financialRequirements to participate should also strike a indicators provided in the bidding documents of the furm to bebalance between quantity (as many as possible) regulated and one of the benchmarks listed below can revealand quality (serious investors) of the offers. They abuses of the regulatory system by the potential operator.should also strike a reasonable balance between Compare the rate of return (ROR) implicit in the bid and:

risks and return for the private investor as 1. the ROR in unregulated firms in similar activities in the

discussed in Box 1.1 region.;2. the ROR the average ROR of the market in Argentina; and

.10 The second prong of this strategy focused 3. the ROR in similarly regulated firms in other provinces

on the nature and form of the dissemination of Over time, the comparison of the variance of rates ofproperty. In the case of Aguas Argentinas, shares return in the sector before and after the inclusion of the new firmwere sold on the stock market to widen the can also provide useful insights on the potential financial impactof the privatization. Typically, utilities are expected to be belowdissemination of property and hence of average risk and their return should vary by less than the marketaccountability. For electricity transmission, shares average. If the bids show anything else, it may reveal that thewere allocated to all staff, giving an incentive to bidder is asking for too much to provide the service. It may also

reflect an attempt b the bidder to compensate for a provinceact in the company's interest. But there are also specific risk that is well above market average. In this case, therisks in spreading accountability too thinly. This specific form of regulation may need to be tuned up as discussedis why it is generally suggested to allow the later in the chapter

bidding of packages of shares large enough toallow control by one major interest (i.e., to have astrategic investor).

4 While reducing the problems due to monopolies, thisapproach still implies a need for regulation.

Table 1.2 Summary of Main Concession Contracts

- Bidding Procedure Term Contract Requirement Rate Regulation Quality Requirements InvestmentWater & Sanitation * technical pre-qualification by OSN * 30 years 0 obligation to provide the public 0 cost plus; 0 water quality * include(Aguas Argenninas for and the Privatization Commission; based on * could be extended by I y service of water and s nitation in a * inal tiff level leves are pelled out, improvement andthe Greater Buenos Aires economic capachy (billing no less than $600 way tht ensures the continuity. at in the prvatization icreingy dem expion pls u paiArea) mnillion/year and net assets no lower than SI * at the endrofte concession, anewbid regularity, quality and generality to procem nd applied to a over tine ofthe contract

billion)andtechnical capacity(toatendin couIdbeorganized usr andtheprotcti onofthe specifictariffstndcure * servicquality * invetmentahaveurban centeri of more than 500,000 nd environment; * trigger rule for target (minimum to be bidden out;total population of 2,500,000); * must extend, maintain and changes based on an Icvels) but no quality * timing

a then, two envelopes: envelope one renew as needed the external network agreed cost structure. norms with respect to requirement for theincludes detailed technical offes (legal and connect them to all inhabited service cuts or water investment program, butaspects of the bidders, mission statement, buildings in the concession zone. * reviisons every 5 pressure levels subject to renegotiationaspects afthe biddem, miuion daten-C ~ ~~~~~~~~~~~~~~~years to increaseoperational plans, regulation for users, ...); * will have to comply with incentive to efficiency * quality levels forenvelope two includes the financia and specific minimum coverage in time treated sewage and foreconomic offers (including the coefficient of slices, ending with 100% by the end sewerage infrastructureadjustment to the current tariffto be of the conitract * fines for non-expressed in value--this is the coefficicnt bywhich the current tariff will be multiplied to * maintain and rehabilitate complvane r ithobtain the new tariff, financial commitment according to needs and minimum revenue returnd toand aptitude and all the information criteria spelled out in contracts usersexplaining how the bidder would achieve the 0 meet demand of both of andadjustment coefficient); new users

* the winner of the bid is the one with 0 meet increasingly demandinglowest adjustment coefficienL quality standards

* beforc bid, create a company with a 0 do secondary treatment of allminimum capital of S 20 m., get guarantee sewagefor S150, niillioii and various insurance; ___ ___

Electricity Distribution * two envelopcs: envelope I specifies * concession for 95 years for exclusivity * obligation ol public service of * RPI - X f- Y; * mininium quality * no conitrol ofin Greater Buenos Aircs die technfical requirements anud detemiincs in a specific zone aller whici new bidding distribution in the concession zone; * maximum price standards covering: investment by the public

the prequalified bidders; it contains the share will have to be made; proceeds of bid go to * commit to specific quality with total pass-through (i) product (tension); sector althoughof actions to be subscribed by each co- the incumbent who is also allowed to bid; levels of the costs of energy in (ii) technical service investments over USS2bidders, shows a single unified local * the term can be extended for a pefiod , respect the rights of users as the wholesale market (duration and frequency million need ENRE'sresidence, proof of legal existence; S10 up to ten years to be determined by the specified in the "reglamento de (Y) and indexation to of outages); and approval and a public

million guarantee for offer, proposed action (iii) commercial service hearing.onunrcialservce heringpillion guarantee for offer, proposed aclion regulatory entity (ENRE) who can also suministi-o,. U. S. price index (RPI); (complaints by clients,plan for contractual obligations Envelop 2 modify or suppress the zonal exclusivity;has the economic offer with amount in cash * satisfy th total demand for * the index used in time to get connection,and public bonds; bidders had to, have assets 0 the term of the concession is divided service, including demand for new 67% PPI and 33% CPI; bill estimates);of at least $300 million and proven net into management periods of 10 years services; * initially X was set a for measurement,wealth of no less than $200 million which (except for the first,. lasting for I 5 years); 0 poieeeg o ulc 1 ;toseswr eie

did not decline by more than 5 to 10% over * at the end of each management period, lighting; * the RPI is applied with different goals;the previous year. ENRE organizes an international bidding to poiecctctyatoascfctrff (i) durirng the first 36* all bids must identify at least one but sell n ajority bundles ofstock and will set 3X380p220 Ve 13t 2iV, 33kV, *rtose tRi months, measurement of

no oretha tw exerince opratrs the tariffregime tobe applied for the 3~o20,1,k,3V tutr;medium tension; and (ii)belonging to the consortium created with following five years; the conditions of0these 132kVd 22kh oR * tariffs are set in as of 37th monthspecific requirements of experience for the biddings have to be similar to the original agreed with ENRE; USS measurement of everyGperator. bidding conditions; the ownier of the 0 made the invcstment and user;,

majority package; if the price offered by the niaintenance required to achieve* the economic offer has to be valid for incunmbent is the highest, he keeps the established quality requirements; * failures to meet

at leat t godays s fromthe oening f othequality standards areat evle 180 days as from he opening of the propety; if he is outbidden, the highest * allow non-discriminatory penalized through aenvelope and at least S30 million would bidder has to pay the bid price to the access to the grid to third parties as dpetailed fin systeru thehave to be paid in cash after 3 days of the incumbent and becomes the new majority long as it does not impede its own proceed to te ewinner selection; 90 days for public bonds. owner, an independent inspector nominated capacity to deliver, victim of the gaps

* the winner would have to pay the by ENRE will ensure I year before the end v of the gapsconsulting fees for the lawyers (USS 1.5 of the management period and for the first ensure that their activities domillion each for the transmission and the year of the ne w e not damage ecosystems;distribution deals) and the finacial proper fiusctioning of the company * promote rational use of energy.consultants (paid only if there is a deal and (including disclosure of information).equal 1% ofthe value ofthe deal).

5

What Contract-Based Regulation Entails contract by redefining one of the variables (thecapacity price) in such a way that the distributors

1.12 The concession contract is the main would no longer be penalized (or favored). Thisinstrument used by the national Government to left formula in a range consistent with the "passregulate utilities--although its exact nature varies through ranges" and avoided the original concern.across sectors as seen in Box 1.2. Concessioncontracts and possibly management in some of the 1.14 Should contract revisions be allowed?poorest regions, are likely to be the main This (as well as the national experience withinstrument for the provinces as well. As a rule as transport contracts) shows that even if, as a rule,mentioned before, they should be drawn up as contracts should be prepared as documents thattightly as possible to reduce the need for are not intended to be modified and includediscretionary adjustments. adjustment mechanisms to avoid the costs,

uncertainty of ex-post negotiation, under veryspecific and very limited circumstances, a limited

Box 1.2 Comparing Concession Contracts in W&S degree of contract flexibility may be good public

and in Electricity, policy. But to ensure that the contract is aThe term concession is used for both the W&S and for credible regulatory instrument, the contract

the electricity distribution contracts. Yet these contracts modifications should be based on some fair andare quite different. No shares were sold in the W&S case.All assets remain public and the Government has given the workable renegotiation rules based on clearlyright to a private firm to operate these assets in exchange spelled out policy criteria (as discussed below)for certain obligations in terms of investment which is one and agreed upon by all involved parties. Theyof the main reason why provincial water utilities are trying should always respect the original contractualto attract the private sector. In electricity, the Government rights of the investors; ad-hoc solutions are notdid sell shares, and hence part of its assets, but has . .cattached to it public service obligations--expressed in terms always i the best iterest of all the partiesof service coverage and quality-- instead of specific involved. This is one of the main rationale for aninvestment requirements. independent regulator who can have some

discretion in the implementation of the rules as anhonest broker.

1.13 However, these contracts, as any othertype of contract, cannot anticipate all exceptions 1.15 When should contract revisions beor qualifications and that corrections are allowed? The challenge is to find a transparentsometimes needed later. Argentina's National mechanism for modifying specific terms of aelectricity regulator (ENRE) was recently contract that do not result in private investors orconfronted with the need to revise a distribution consumers lacking confidence in the contractualcontract. The formulas for the calculation of instrument (i.e., adjusting the price terms of atariffs prevailing initially resulted in a significant contract in line with the terms of that contract).reduction in the profits of distributors in May This mechanism was missing in the otherwise very1994 (although the mistake led to significant impressive national reforms. It is also a keyprofits over the previous 18 months without much ingredient missing from the provincial debate.complaints about this problem then). Thedistributors complained about their profit losses 1.16 The contractual rigidities built into theand ENRE conceded that there was a need to concession agreements are necessary to closerevise the contract but that this required some deals and to create binding commitments amongresearch. The Energy Secretariat (ES) took the participants. However, they make it difficultcharge of the debate and de facto revised the to adapt when there is a need to resolve emerging

6

problems because many of the actors find 1.18 Which public policy criteria? Theadaptation threatening to the privatization public policy criteria to test if a revision is neededcommitments that protect their interests and the have to be spelled out as soon as possible to makewhole fabric of reform. Moreover, clearly each the rules of the game clear to all parties involved.franchisee is likely to attempt to interpret most of They should ensure a transparent basis forthe contractual ambiguities to its own advantage. contractual dispute resolution and related policyThis is why there is a need to continuously decisions, as well as to avoid excessivelymonitor the concession agreements and to assess subjective decisions on the need for or nature ofany need to adjust them. contractual changes. These criteria ease the

judgment as to whether the maintenance of1.17 The assessment of any modification existing provisions is suboptimal to all parties inrequirements, however, needs to be based on a view of a fundamental change in externalgood sense of what went wrong and on a clear set circumstances (e.g., a permanent demand shift) orof public policy criteria: a change in policy priority (e.g, the relative

.How realistic are the government requests importance of services to the poor increases)

once more is known about the state of assets? Possible criteria include:Were there trade-offs not well identified at the * Is the protection of the interests of investors atinitial stage by either the Government or the the baseline levels established in the originalprivate bidders who could have asked then for privatization terms guaranteed'? Should it be?a revision of the specification of the needs, as .How would the overall operation (flexibility,observed in the privatization of the Santa Fe

water company. ~~~~~~variety, responsiveness of operators, quality ofomaintenance and of investment strategies) be

* If minimum demand levels were spelled out in affected by the changes?

the contracts, were they over- or * Would the composition of the financing of theunderestimated? This is common. It may activitybe altered? Would it reduce the publichave been a problem in any of the waterconcessions in Argentina as elsewhere in the share in this financing?world. But was this due to a mistake in the * Would consumer interests be protected?organization of the concession or was it due to What is the nature and source of change ina mistake by the private operator such as scale of operations in response to demand?insufficient market analysis?

* Did the state of the assets deteriorate between If the answers warrant a change to the contract, itthe time the bids were made and the time they should be limited to the specific issue.were actually transferred? If yes, was this due Negotiations should not be reopened for theto the natural phenomena (e.g., the weather) whole contract.or to the fact that the public operators stoppedmaintaining these assets? If it is due to the Does the Provincial Choice of Regulatoryformer, the responsibility may be shared by the Regimes Matter to a Potential Investor?Government and the private concessionaire.In the second case, the concessionaire may 1.19 The two main regulatory optionshave a fair claim on a request to revise the provincial reformers should be picking from are:contract. (i) rate of return (ROR) regulation (as used for

over 20 years the U.S. utilities) and (ii) some formof price cap regulation (recently introduced in the

7

UK privatizations and in the electricity distribution What Decentralized Regulation Entailsin Greater B). Under a ROR regulation, theregulator set a revenue requirement based on a 1.22 The last provincial challenge addressedfirm's accounting costs reflecting operating costs, here is the organization of the monitoring of thetaxes, amortization and allowed ROR. Once the consistency of the behavior of the privatizedrevenue requirement is computed, the regulator monopolies with the public aims. This monitoringdetermines the tariff structure design needed to is needed because concessionaires will have strongrecover aggregate costs. Under a price cap, incentives to behave inefficiently when leftinstead of setting a ROR, the regulation sets a unsupervised. The creation of regulatory bodiesprice ceiling above which the concessionaire or the use of the strength of law and courts will becannot raise prices. Under that cap or ceiling the an important element in the success of theregulated firm can set prices as it wishes. The provincial reform of utilities. They are needed toceiling must be reviewed every two to five years guarantee the increase in efficiency throughto account to productivity improvements, competition in services delivered by natural

monopolies.5 But there is a limit as to how much1.20 Price capping has many potential this institutional role can and should beadvantages over ROR regulation. The main decentralized.advantage for Argentina's provinces is that itgenerates stronger incentives to cut costs and that 1.23 The main economic argument in favor ofthese costs cut eventually get passed on to some degree of decentralization stems from theconsumers. One of its problems is its very need to distinguish between technical anddemanding informational requirements. This can economic regulation. An interconnected electricbe overcome as ENRE's experience is showing network will always need technical regulationalthough it is not straightforward as shown by the with strong coordination between the differentBritish experience. But the main reason why regulating agents, for example through a centralprovincial reformers have to be careful in their dispatching unit. Economic regulation does notchoice is that the specific choice has an impact on have the same goals, so that its organization is notthe risks faced by potential private investors. The necessarily identical to the technical one. Thetwo regimes place very different levels of risks on controllers should be located at the nodes wherethe regulated utilities and hence affect the rate of inefficiencies will more probably occur and thesereturn and the cost of capital in very different nodes may differ for the technical and theways. economic concems.

1.21 A price cap approach implies that the 1.24 Economic Regulation of Electricity.investors puts up with all the risks on its The technical dispatching of energy supply toinvestment. ROR can pass on these risk onto the meet demand is a good opportunity to induceconsumers. So if demand for utility service in any efficiency through merit order. On the contrary,province is highly dependent on the level and type at the terminal nodes of the network there is littleof economic activity in the province, ROR need for technical control but it is a place whereregulation will generally be more effective at distributors can try to exploit their exclusiveprotecting the investor. In other words, high risk position in front of captive consumers.provinces should consider allowing rate of return Consequently, economic regulation is necessary atregulation when it is not finding any privateinvestor interested in providing the services under 5 This has to be supported by a strong commitment toa price cap regime. antitrust but this is a commitment that has to be made

at the national level, not by the provinces.

8

the regional level. But this means neither that one with a single independent entity for all utilitiesagency is necessary in each province nor that the with a modest staffing level dependent on theprovincial regulators are to be independent from requirements of specific tasks assigned to thethe central agency (e.g., ENRE). regulatory authority rather than multiple sector

specific agencies. The main reason is that the1.25 In view of the similarity of problems in skills needed to be a good regulator are hard toseveral provinces, a small number of three or four find. Luckily, the skills needed to be an effectiveinter-provincial agencies should be sufficient. If regulator are very similar in both the water andthe political situation is such that each province electricity sector and hence the same resourcesmust have its own regulatory body, the entity can be shared within a single provincial regulatoryshould be kept to a minimum. Moreover, entity. This means lower resource needs (not onlywhatever their number these local agencies should in terms of staff but also in terms of equipment)have only specific prerogatives, essentially and hence a lower burden on the taxpayers or theauditing and reporting to the national regulation beneficiaries of the services. There are otherbody. This is because technical dependencies benefits. For instance, it facilitates learningwithin an electric network are particularly strong. between sectors, ensures consistent approaches to

central policy questions and makes the regulator1.26 Economic Regulation of W&S. In less vulnerable to industry or political capture.W&S, things are somewhat different since water This lower vulnerability enhances the credibility ofcannot be collected anywhere, neither dispatched the agency and reduces risks for investors.in any direction independently of geographicconsiderations unlike electricity. Water networks 1.28 The implementation of this strategy can beare not national. But this does not imply that each flexible. It is easier to do when agencies have notlocal network in each municipality should be yet been established. A multisectoral frameworkregulated by an independent body, because there can be established first and then sectors can beare important economies of scope in the added to the core entity when appropriate. Thereregulation of pumping, treating and distributing is no need to wait until any contract has beenwater at the local level. The informational, signed. The creation entity can be instrumental intechnical and organizational problems are assisting in the implementation of theisomorphic from one municipality or province to concessioning process. If one agency has alreadythe other. Consequently, the optimal organization been established, it should not be too difficult tofor the regulation of this sector should include a expand the jurisdiction of the existing agency.national entity in charge of the definition of This depends of course on how sector specific ingeneral principles like pricing rules, quality design and composition the initial agency is.standards, uniform statistics (for yardstick Finally, if several agencies have already beencompetition) and provincial entities controlling created, the best option is to develop a strategyand overseeing the application of these rules. for merging the agencies after an initial period.

This strategy tends to be the most difficult and1.27 The institutional dimension of will often be resisted by both the existingprovincial regulation. If most provinces regulators (concerned about losing their autonomydeciding to rely on concession contracts for the and possibly their job) and investors (oftendelivery of electricity and W&S service end up preferring the counterpart they know).deciding to create their own regulatory entities, afew words of caution are needed. Chapter 4shows that most provinces would be better off

9

Chapter 2: REFORMING ELECTRICITY DISTRIBUTION IN THE PROVINCES

2.1 The main purpose of the reform of of this report. To be able to identify the optionsArgentina's electricity sector was to reach for reform in the provinces, however, anefficient pricing and production levels in the short- assessment of the achievements of the Nationalterm, and an investment level sufficient to meet program so far is needed. This is why this chapterdemand in the long run. This entailed a major starts with a general overview of the sector as itrestructuring of the sector which started with the stands after the main national reforms. Next itlegal initiative expressing intentions in 1989, was discusses the main institutions that could influencefollowed by the first implementation steps in 1992 the regulatory environment of privatizedand is still going on. provincial distribution companies. It makes it

clear that the provincial regulatory authorities are2.2 The legal basis of the restructuring process not going to be the only institution that will have ais spelled out in the 1989 laws deciding the global strong impact on the return to investment in thereform of the state. For the electricity sector, the provinces. This is also clear in the discussion ofprocess began when the federal government the main regulatory instrument--the contract--andfranchised the distribution and commercialization of the mechanisms of competition. The chapteractivities of SEGBA,6 the vertically integrated concludes with a discussion of pricing inutility supplying electricity to 15 million people in generation, transmission and distribution as thethe Greater Buenos Aires area. The main next rate of return of the private investors in provincialstep was in 1992, with the privatization of the companies will be influenced by the full pricingelectric generation and transmission activities that chain. The chapter concludes with a discussion ofSEGBA was still carrying.7 With these two the importance of an explicit linkage betweenchanges, the original public firm had been tariff design and investment needs in distribution

8vertically disintegrated into seven business units: services.four generation firms, and three distribution firms.These units were either sold or concessioned to Overview of the Sector's Organizationthe private sector through international bids. Thereform in two other state-owned entities, Agua y 2.4 An effective way of visualizing the extentEnergia Elctrica (AyE) and Hidronor, had some of reform in the sector is to follow the financialimplications for a few provinces as some of the flows. Depending on whether a specific activityassets were privatized while the plants located in between generators, transmitters, distributors andareas under provincial responsibility were users is done in a competitive market or not, thetransferred to the provinces concerned. related transactions are regulated in different

ways, as Figure 2.1 illustrates. It shows that2.3 The remaining step in the restructuring of contracts between distributors--and large users--the sector is the reform of the provincial and generators are not regulated and that the spotdistribution companies which is the main interest prices and seasonal prices are also set by the

market. Final users tariffs are, however, regulated

6 Servicios Electricos del Gran Buenos Aires. which should matter to investors in provincialLaw No. 24.065 (December 1991) and DecreeNo. 1.398/92 (January 1992) establishing the 8 Most of the background on the description of the"Electricity Regulatory Framework". In January 1991, sector presented in this chapter is from Bastos, C.M.SEGBA had a generation capacity of 2500 MW and A.A. Abdala (1993), Transformaci6n del sectorsupplying 10.33 TWh. to 4.5m connected customers. electrico argentino, Editorial Antartica, Chile.

10

utilities. So is transmission price and this is 2.7 The market matches electricity demandimportant to these investors since depending on and supply with an hourly price.10 The marketthe effectiveness of this regulation, transmission also allows trade-in contracts, in which suppliersexpansion will be responsive to the distributors' and buyers can freely agree on long-termneeds or will not. The chapter shows how and contracts in quantities and prices as well as morewhy these activities work the way they do and qualitative aspects such as voltage, point ofwhy it matters to the provinces. reception, timing, back-up, etc. But these do not

imply any reranking in the order of dispatch as2.5 The discussion starts with a brief overview discussed below."of the characteristics of each one of the mainactivities, then moves on to discuss the main 2.8 The co-existence of three different types of"regulators" of the sector (in a broad sense). players on the demand side is important toNext, the chapter explains how contracts have recognize as it matters to the value of the assets ofbecome a key regulatory instrument for these a distribution company and for the design of theirfederal regulators and how competition works in tariff policy as discussed later. These playersthat regulated environment. Because pricing is include distributors and (potential) foreign buyerssuch a key component of the effectiveness of the but also large users. Defined as those whose peakreform and of the regulatory function, it is demand is equal or higher than 0. IMW, largediscussed next in some detail. This discussion users have the advantage over other consumers:spells out all the payments for service obligations they are allowed to sign direct contracts withas these are likely to be accepted and internalized generators without having to commercially goby provincial governments in their own reforms. through provincial distributors and henceThe discussion also addresses the linkages increasing the incentive to minimize costs for thebetween tariff design and incentives to invest as operation of distributions services. They are ablethis is, or at least should be, a major concern of to access the MEM directly for at least 50% ofmost provincial governments. It highlights the their total demand.potential consequences for the provinces of failingto address the linkages between tariff design andtransmission expansion.

2.6 Generation. The core of the reform ingeneration was the creation of a wholesale spotmarket, MEM. The provincial companies aremajor actors in this market as it is open to anygenerator--whatever its technology--and the mainusers (distribution companies but also deregulatedlarge users). These can directly buy from anyprovider they chose to on that market9 (seeBox 2. 1).

'° Real time tariff sets the price of the service at anytime, reacts continuously to changes in demand andequalizes demand with capacity at all times.Besides the MEM, there are two other small wholesaleelectricity markets which are not connected to themain transmission system (Misiones and SouthPatagonia). Prices in these markets will be determined

9 Mercado Electrico Mayorista. locally until they are connected with the MEM.

11

Figure 2.1 Financial Flows

G2 X Wholesale spotmarket

Transmission Co| Un

Gi: GeneratorsDi: DistributorsUi: Users

Contracts. Unregulated pricesSpot price. Marginal cost of last dispatched generatorSeasonal price. Forecasted average of spot pricesFinal users tariff Regulated tariff with pass-through of seasonal pricesTransmission tariff Regulated tariff

12

Government made the most of the possibilities ofBox 2.1 Who are the Main Generators? benefitting from market-based regulation by

The creation of the MEM resulted in an increase in the organizing an auction for the right to deliver thenumber of generator from 10 before pnvatization to over 30 transmission services but also by building inindividual companies. The five largest individual generators are periodic threats to replace the concessionaire withCNEA Central Nuicleo E1kctnca Argentina, the state-owned a challenger. This serves to oblige the incumbentnuclear generator and the privatized Central Puerto (thermal), to behave efficiently once the contract has beenCentral Costanera (thermal), ESEBA (thennal) and CTM SaltoGrnde (a binational hydro generator). With the privatized awarded and guarantees the minimization ofHidronor andAguay Energia Eletnca (almost fully privatized), transmission cost to distributors and generators.they supply about 75% of the electricity needs.

In 1993, about 46% of the production was thennal, 2.13 The exclusive concession for transmission37% hydro, 15% nuclear and 2% imported. Salto Grande and iS for 95 years. This total duration is, however,Comahue are the main hydro sources. Most thermal generators divided into a sequence of management periods.are located in the Bs.As. province, the main center of demand. The first is for 15 years and is followed by I 0-year

management periods. Six months before the endof each management period, the sale of the

2.9 On the supply side, any firm wishing to concessionaires' controlling block is organized byproduce electricity can enter the market if it ENRE (jointly with the tariff regime to be appliedrespects technical and safety standards for its over the following five years). The concessionairenetwork connections. The main actors are is also one of the participants in the bids. If itsindependent power producers, national offer is lower than any other, the concession isgovernment generators, binational utilities and lost and the government reimburses the departingf>oreign producers. concessionaire the value of the sale (net of debt).

2.10 Firms already operating a distribution 2.14 Distribution. There are 22 mainnetwork are not allowed to own generation to distribution companies across Argentina movingprevent new vertical integrations. To protect electricity from the transmission lines to the localdistributors and generators, however, carriers are customers. The three largest are Edenor, Edesurnot allowed to buy or sell power even if electricity and Eseba in the Bs.As. province and they buyis physically received by them. This exclusion is almost 60% of Argentina's consumption. In someneeded because transmission is a natural provinces (e.g., C6rdoba, Chubut, La Pampa andmonopoly and monopsony and carriers would end Neuquen and in the interior of the province ofup enjoying excessive market power. Bs.As.), cooperatives can be major actors in the

sector, but these are not addressed here.2.11 Transmission. The transmission activity,in contrast with generation, is considered a natural 2.15 From a regulatory perspective, there is nomonopoly and costs are minimized when there is major difference between high voltage nationalonly one firm delivering the service in a given transmission and low voltage local distribution.area. This is why the government of Argentina is, Both are natural monopolies. Given the federalcorrectly, carefully monitoring the decision of the jurisdiction of the RMBA (metropolitan region oftransmission companies to prevent them from Bs.As.), Edenor, Edesur and Edelap are regulatedinefficient behavior (see Box 2.2 on the sector's by the National Government. All the otherorganization). distribution companies are or will have to be

regulated by the provincial governments. Since2.12 Although competition in the operation of a the national experience with distribution servicesgiven network would be inefficient, the National

13

is clearly the model being followed by the potential investors as one of the main risks. Thisprovinces, a brief overview of the rights and is why a commitment to a strict enforcement ofobligation may be helpful. property rights is key to the success of the

attempts to privatize (see Box 2.3). Because ofBox 2.2 Who Transmits Where in Argentina? the vertical disintegration, distributors also have

At the national level, Transener is the main player and to buy electricity on the spot market or signowns and operates all of the 500kV and some of the 220kV contracts with generators. This is the competitivetransmission lines. It is a private firm with a 65% control by a side of their activity. The remaining govenmentconsortium of seven almost equal owners (the government kept25% of the shares to be sold on the market and the employees role is to regulate these monopolies.received 10% of the shares). The winning consortiumn proposed a

US$ 234 millions bid. The losers' bids were respectively 2.17 The main legal obligations of the formerUS$l7lm, US$1 52m and US$60m which shows the large feraditbuoncm nesw thrpvtzediscrepancy in the valuation of the company despite the open federal distribution companies whether privatizedinformation concerning the assets. or not are:

Transener links the main areas of generation (Neuquen, * obligation to meet all the present demand andMendoza, Tucuman, El Chaco and Entre Rios), with BuenosAires where most of the demand is based. The network being any increase of demand;radial and with few interconnections, a failure in any of the three * complete freedom to determine its investmentmain line can disrupt the supply in Bs.As. The construction of afourth main line is between Comahue and Bs.As. is under program as long as it satisfies predeterminedconsideration. levels of quality for technical product, for

hi the south, Transener is essentially transporting huge technical service and for commercial service;quantities from the production area of Cornahue to theconsumption market through three long-distance lines. In the * obligation to grant open access to the networknorth, it interconnects regions that were formerly supplied by to any generator and customer; andtheir own generation plants. Transener is also responsible for thequality of service provided by independent carriers and has to * have a legal residence in its concession area.supervise the building and operation of their installations. It is

compensated for this supervision. Provincial distributors can have other obligations.There are also five regional transmission companies: The obligation to meet ALL demand might not be

Transnoa, Transpa, Transnea, Transcuyo and Transcomahue.All except for the last one have been privatized. All these realistic in sparsely populated provinces withouttransmission companies are organized into the Argentine System subsidies.of Interconnection (SADI) which covers about 90% of the

country. 2.18 The exclusive concession for distributors

2.16 For companies under federal jurisdiction, in Greater Buenos Aires is for 95 years. This totaleach firm receives an exclusive concession to duration is however divided into a sequence ofdistribute electricity in a specific geographical management periods. The first is for 15 years andzone but is unauthorized to manage any other is followed by six 10-year periods. Six monthsactivity.'2 This a perfect illustration of what a before the end of each management period, theclactivity. ent isfa p.ert iustraio lo waa sale of the concessionaire's share is organized byclear assignment of property rights should do. .The lack of commitment of some provincial ENRE (jointly with the tariff regime to be appliedgovernments to such a clear respect of private over the following five years). The concessionairerights, not without occasional attempts to violate is also one of the participants in the bids. If itsthose rights--for instance by failing to pay bills due offer is lower than any other, the concession isthose rubigh ets or instrce by farceiing etodp bil dy lost and the government reimburses the departing

concessionaire the value of the sale (net of debt).

'2 In C6rdoba or Mendoza, for instance, generation anddistribution are held by the same company.

14

The "Regulators" Box 2.3 What did the Concessionaires do withtheir Property Rights in Distribution?

2.19 Regulatory objectives. The objectives of The distribution part of SEGBA, one of theArgentina's national regulation are to: companies under national jurisdiction, was divided into

three companies, Edenor, Edesur and Edelap. They* benefit from the strength of competition in represent 47% of Argentina's market. Their problems

promoting static and dynamic efficiencies were typical of those of distribution companies in the 53%while protecting consumers from undue managed by the provinces: poor plant equipmentappropriation of gains by firms; condition, low productivity and high levels of theft. Over

the last two years, for instance, customer "losses" inorganize cooperation between firms when Edenor and Edesur amounted respectively to 30% andthere are externalities beneficial for social 25%. Yet they are only allowed to pass through about 11%welfare without encouraging collusion; and of the costs represented by the losses to consumers.

To cut these losses, Edesur reduced technical and* redistribute risks and gains according to other losses (e.g., theft), and employment and overtime.

political and social guidelines without Employment was cut from 7,417 in January 1992 to 4,677excessive distortions in the economic in October 1994. The cut in Edenor was from 6,368 toindicators of scarcity. 3,759. Overtime allowed was reduced from about 40 hours

per employee per month to about 3 to 5 hours.These are likely to be the main concerns of The reduction of theft was a more complexprovincial regulators as well in this sector but they activity because it involved shanty towns and low incomewill have to work with various other partners in areas where cutting power off was not viewed as acceptablethe implementation of these goals. by the local authorities. The solution was to negotiate a

deal with the Regional Metropolitan Area. Edenor andEdesur agreed to place a medium-tension distribution

2.20 The actors. The main public actors network in those areas and connect only ten or soinvolved include: consumers with low-voltage lines and related transformer.

The voltage in these medium tension lines is too high for* The National Energy Secretariat (ES) illegal connector. Any illegal use of the low voltage line

will be easy to identify as the transformers installed will* The National Regulatory Entity (ENRE) not be able to process more than ten users. This places the* CAMMESA burden on the entire commuunity to solve the "illegal"

problem.* The Provincial Regulatory Authorities polm The bills are paid to the concessionaire by the

municipality who takes on the responsibility of collectingfrom the users. The outcome has been obvious. Payment

2.21 The Energy Secretariat. The ES is under levels in low income areas have reached 90% for Edesurthe jurisdiction of the Ministry of Economy and while Edenor has half of the low income population payinghas overall responsibility for the industry as its bills.defined in the framework set by Congress and theregulatory decree set by the government. Its main 2.22 ENRE. ENRE is the autonomous nationalrole is to set sectoral policies. It defines the main regulatory agency. The government designatesoperating and developing rules, grants and the 5 members of the board with the agreement ofcontrols national concessions and has the final say the Congress but two are proposed by the Federalin sector disputes. Energy Council. Its financial resources originates

from the agents that participate in the wholesalemarket; they pay in proportion to the volume oftheir interventions (see Table 2.1 for moredetails).

15

Table 2.1 Structure, Financing and Functions of ENRE, the National Entity for the Regulation of Electricity

Structure * Board of five directors appointed by the government* staffing for the president and vice president based on a public advertisement (requirementsincluded being an engineer or an economist and specific experience; the final selection was basedon interviews conducted by a private consulting firm--Price Waterhouse-- who short-listed of threefrom which the Energy Secretary selected the President and the Vice President); one of the directors,was solicited based on public job listing (with conditions including economist, engineer or lawyerqualifications, following the same procedures by the private consultant) and the last two were basedon the short list proposed by the Federal Energy Council (which is an organism of the provinces).

Financing * makes and publishes its own budget annually, to give a chance to all agents to object to it.* budget has to be approved by parliament (as part of the national budget)* has a staff of 75 of which 20 deal with complaints by users* the budget resources are from:

(i) inspection and control fee: producers, transporters and distributors pay anticipativelyan annual fee set differently for each agent. Each agent pays in proportion to his/her ownshare in the value of gross production of the wholesale market(ii) subsidies, inheritances, donations of any type(iii) any other legally mandated source of revenue(iv) fines(v) interest on own resources investments

Functions and * enforce the regulatory framework, contracts and public service obligationsObligations of the entity * issue rules and regulations on matters of safety, technical procedures and norms, measurement,as spelled out legally billing, control and use of meters, interruption and reconnection of service, access and service

quality* prevent anticompetitive, monopolistic, discriminatory behavior at each stage of the process* define the basis for the calculation of tariffs set in contracts* control tariff enforced* publish the general principles to be respected by distributors and transporters to ensure free access

to their services* determine the basis and criteria for the assignment of concessions* organize and implement the bidding, adjudication and signature of contracts* organize public hearings* monitor the respect of property rights, environments and public safety* take to court the relevant issues* regulate the proceedings to impose sanctions* impose sanctions* publish information for and advise generators* issue an annual report and recommend actions to the executive power when needed* do whatever is necessary to ensure respect of the law* collect information from transmission enterprises

2.23 ENRE ensures the compliance with the the tariff calculations, which affect investmentElectricity Act. It supervises and regulates the plans. It interprets the rules defined by the ES andsector, ensuring that ES's strategy is enforces compliance through penalties, suspensionimplemented."3 But more importantly from the of operations, terminations of concessions andprovinces perspective, at the transmission and other measures. It limits the opportunisticdistribution levels, ENRE establishes the basis for behavior of the network operator that could take

advantage of its double monopoly position(upstream and downstream) to extract too much

13 Ente Nacional Regulador de la Electricidad.

16

profits. It checks that its pricing policy conveys operations through the transmission network.i6 Itaccurate scarcity signals from generators to final also performs settlements for all participants in theusers without distortion. It is also responsible for MEM (i.e., collects transmission revenues fromdefining the criteria and conditions for awarding the users and distribute them to the carriers).new concessions--jointly with ES-- and for issuingall regulations regarding safety, standards and 2.27 Its dispatch decisions aim at minimizing theprocedures for technical areas such as metering generation and transmission costs while satisfyingand service quality. demand. This role as network controller is

complicated by the geographical dispersion of2.24 ENRE also has settlement responsibilities. generation sources and consumption destinationsIt has jurisdiction over all disputes relating to and by the high variability of the timing of plants'supply, distribution and transmission. ENRE availability and users needs. This means thatsettles conflicts between CAMMESA and an CAMMESA is a large consumer of technical andagent. It has the power to apply penalties for non- economic information. Because of its presence,compliance and initiate and pursue legal action to the individual firms that sell and buy on the M:EMensure compliance. have to take account of the various externalities

they create.2.25 CAMMESA. The main purpose ofCAMMESA is to coordinate supply and demand.14 2.28 The Provincial Regulators. TheCAMMESA is Argentina's solution to the only provinces are already present in the regulatorypotential issue in its market-based regulation of framework through the Federal Council of Electricgeneration. This issue stems from the difficulty of Energy (CFE). This council is made up ofinstantaneously adjusting supply and demand in an representatives of each province. It has aninterconnected network. The adjustment requires advisory capacity (to the ES) and ensuresa dispatching activity since demand is not perfectly consistency and coordination in regional policiesknown until the very last moment. 15 for the sector. In the near future, CFE is likely to

be much more actively involved as more and more2.26 CAMMESA is a joint-stock company-- provinces manage to concession their distributionnonprofit oriented--owned in equal proportions by companies. A provincial regulatory entity hasthe ES and associations of generators, distributors, already been created in eleven provinces. Thetransmission carriers and large users. Widely privatization process may also require some formspeaking, it implements the operating rules issued of redefinition of the role of the Federal Board.by ES. CAMMESA's main task is to control the This took place recently with the formalization ofexchanges in the bulk power market and the the coordination of provincial sectoral regulations

at the national level, signed on April 24, 1996 bythese provinces. Many provinces are still

4 Compailia Administradora del Mercado Mayorista considering a local regulatory or supervisory role.Electrico SociedadAn6nima. But they will, with ENRE, undertake many of their

responsibilities jointly: studies, exchanges ofs Another reason is that the evaluation of water in experiences, training of their staff. organize

hydroelectric plants cannot be performed by market technical assistance, disseminate information torules because of the alternative uses and theenvironmental impact of this resource. Note also that users, coordinate specific regulatory activities.all the state-owned generation plants were notprivatized: nuclear and binational plants that count foralmost 40% of generation and 15% of capacity are stilla property of the state. 16 Sistema Argentino de Interconexi6n (SADI).

17

This should lead to the creation of a "super few provincial companies mentioned in Table 1.1 aregulatory agency for the sector". and 1. Ib). Since generation has become a truly

competitive activity in a service considered of2.29 An implicit key player. Large users are, general interest but organized as a risk- orientedindirectly, important actors. Because they can activity, public assets were sold rather thancommercially bypass the distributors in buying concessioned and the market rather thantheir electricity, they minimize the leverage concession contracts takes care of the regulatoryprovincial regulators can or will have on them. needs. The only obligation for the buyers is toThey end up being mostly overseen by the ENRE accept the technical and commercial ruleswho monitors their contracts on the wholesale governing the functioning of the market if theymarket. 17 The relative importance of this factor in want to join the Argentina Interconnection Systemthe distribution of responsibilities has been to sell their output.increasing since the beginning of the reformprocess, as the definition of large users has tended 2.32 The contracts in all these cases have a veryto become increasingly encompassing through the similar design and also provide a good model forreduction in the definition of peak demand level all the provinces which have not yet organizedneeded to qualify as a large user. The customers their reforms. The concessions are for 95 yearsauthorized to intervene were first defined as those with exclusivity in a specific zone after which newwith demand of 5MW or more; and after bidding will have to be made. The revenue fromsuccessive decreases the threshold was fixed at these bids will go to the incumbent provider of the100 kW on February 1995. service who is also allowed to bid to continue

delivering the service. The term can be extended2.30 There are in fact two types of large users. for a period up to ten years, as determined by theA GUMA'8 is a user who has a demand of at least regulatory entity who can also modify or suppressof IMW and consumes more than 4380 MWh in the zonal exclusivity. A peculiarity of thethe year. Its contract with MEM generators must Argentina model is that the term of the concessionhave a minimum duration of one year. To be a is divided into management periods of 10 yearsGUME, the user needs to have a minimum (except for the first lasting for 15 years). At thedemand of 0.1MW and to sign contracts for at end of each period the industry regulator organizesleast two years. To be allowed to participate in an international public bidding procedure for thethe MEM, large users have to contract at least for sale of the majority block of shares on the50% of their expected demand. If not, they have operating firm and determines the tariff rules forto buy directly from their local distributor. the forthcoming years. The owner of the majority

block at the date of each public sale can participateContracts as the Main Regulatory Tool under specific conditions (sealed bid) and keep the

operating rights for free if it proposes the highest2.31 The main role of concession contracts in bid. If a challenger is the winner, it gets theArgentina's electric sector is in transmission and concession and pays the price reimbursing thedistribution (for now only the distribution former owner for investments already made.companies formerly under national control and the

2.33 In some provinces where the public sector

Moreover, it affects the value of a concession for a has not been an effective manager but where theprivate investor as large users can exclude themselves private sector is not willing to take on afrom the concession market and deal with generators. concession contract because of concerns with theGUMA: Gran Usuario Mayor. GUME: Gran provincial government's commitment to reform,Usuario Menor.

18

short-term management contracts (two to three in the provinces where they are still public but alsoyears) may be an effective way of quickly some of the privatized ones) still have links withcommercializing the sector. Its main drawback is generators. So there is a risk of anti-competitivethat it is unlikely to generate much private behavior and there is not much the provincesfinancing but it should at least reduce operational would be able to do about it without the help ofcosts in the service. If, for whatever reason, the some national regulation. This risk is the mainprovincial government prefers not have any private reason why the federal regulator allows consumerssector involvement, they may wish to consider to buy directly on the bulk market, although itimposing performance contract on its managers limits access to the wholesale market to largeand replace those who cannot deliver the goals customers. It is a good compromise and easy tospelled out in the contract. adjust as the definition of a large user can and has

been adjusted over time.How Competition Works

2.36 This direct bulk market promotes2.34 Some form of competition has been competition between distributors and large users.introduced into the three "functional" stages. For But this does not mean that the physical bypassingdistribution services--as well as transmission--, the of distributor is necessary. In some cases, itallocation of concessions through auctions allows requires the installation of new lines and prior toex ante competition. But these activities are reform, many large users had to build their ownnatural monopolies and conditions for ex post connections to local grids--they also had to paycompetition are not easily satisfied, as discussed discretionary charges to subsidize local (civil)below. In contrast, generation allows a true project. New lines however would be inefficientmarket competition. The reform created 25 from both a private and a public point of view.business units sold separately to private operators. This inefficiency is avoided in Argentina by a lawThey can sell electricity capacity on the spot requiring that the distributor transmit electricity tomarket or by means of contracts. When large large users who purchased it on the wholesalecustomers contract directly with generators they market but allowing this distributor to charge anbypass commercially but not physically their local access fee for the use of the lines.distributors. So, this organization supposes thedefinition of strict rules for the access to the Pricingdistributors' infrastructure and for the fees paid tothe distributor for this service which need to be 2.37 The access fee is one of the many ways inrecognized by the provincial contracts. Gaps or which pricing contributes to the sustainability offailures in these rules can affect significantly the the system and obviously provides an importantvalue of distribution company in the eyes of a financial guarantee to investors in distributionpotential private operator. services. But this is only one of the pricing

aspects that is relevant to the provincial regulators.2.35 By removing the ability of the transmission The other pricing issues are discussed in somecompanies to function as buyers, the risks of its more details below. These issues are reviewed in apotential to exert monopsony powers are general context next.eliminated, but this also creates its own problems:it restricts the list of buyers to distributors mainly. 2 38 Generation Pricing. The price pm of theThis has two drawbacks: first there are relatively distribution companies' supplier is unique, set byfew of them, and second, for historical reasons anddespite vertical disintegration, most (in particular

19

the market and computed at the load center.'9 It is auto-generator is connected to the grid to exportequal to the system's marginal cost of generation its excess capacity. The node factor is then lessmeasured by fuel costs (see Box 2.4 on the than one because the producer has to pay for themeasurement of marginal costs for different types transport costs to the load center and will thus beof generators) This equates demand and supply paid less than the market price. The closer theand guarantees that the price of the last MWh load center, the close the node factor is to one. Atbought is equal to the cost of supplying it. It also importing nodes, electricity is taken from thesignals to users the proper value of their electricity network by distributors or large users. Their loaduse and allows distribution companies to make factor is larger than one because they have to payinformed decisions on this component of their more than the market price to cover thecosts. transmission costs.

2 39 Although not a serious problem, this signal Box 2.4 How are the Marginal Costs Computed?is somewhat biased. There are indeed other The short run marginal cost is determined by thecomponents to short run marginal cost that marginal cost of the most expensive thermal power plantdeserve some attention in view of the wide running to meet demand in any given hour. It isdifferences in technologies across generators. established by CAMMESA for each generator on the basisWhile it is likely to be an unusually good first of its fuel cost and its thermal efficiency. For hydro

sources (about 37% of the total), an opportunity cost ofapproximation of the short run marginal costs, any water is computed with a software allocating prices to theopportunity for revision should consider the substituted fuel. This cost of water is initially used forinclusion of some of the other factors. Some dispatching purposes only. It is then not reflected in theadjustment was introduced through Resolution calculation of the market price. Since ResolutionI 05/95 allowing terierators a band of 15% on fuel SE 105/95, water can now determine the marginal cost.costs t( o allow for other costs. Also hydrogenerators now declare their own schedule andcosts. 2.41 The merit order. CAMMESA dispatches

the energy from generating plants by order of

2 40 In practice, a price 'pn" is set at each node merit, which means by order of increasing shortto remutnerate the generator supplying electricity run marginal cost of production (the cost ofto node n and is paid by users located at the exit generating and transporting to the load center annode. It is determined by pm and a nodal factor. additional kilowatt). The clearing market price isThis nodal factor is a measure of the established generally set by the most expensive generatortransmission losses between the entry or exit point needed to meet demand. Any generator with aand the load center of the MEM. Its calculation is cost lower than the clearing price makes a profit.done in real time with a sophisticated software This profit is the difference between pn (which is amanaged by CAN'IESA. It varies according to measure of the short run marginal cost for thethe season and time of day (peak, off-peak or whole system) and the generator's own short runother hours). It depends also on whether the marginal cost. The existence of this profit pushesnodal point is exporting (<I) or importing generators to cut their costs as much as possible topower (-> I) An exporting node is the node at improve their standing on the merit order.which a generator is connected to the market or an

2.42 Figure 2.2 illustrates the mechanics of the

9 Except if there is a transmission limitation on the system. When demand is D2 , the price is pm -transmission network that does not permit an optimal MC 2 and the firm that operates the plant K1 earnsdispatch. In this case, one defines a local price for thesei of nodes belonging to the restricted area. profits equal to (MC2 - MCI) * K1. If the level of

20

equipment is optimal, this short run marginal cost la potencia puesta a disposici6n, PPAD), atpricing is sufficient to pay all the costs, including present $10/MW, is fixed by the Secretary ofcapacity costs. Energy.2 0 There is also a market for cold reserves

(reserva fria) in which generators bid prices to2.43 The need for a capacity payment. What cover the need for reserve established byhappens when the equipment is not adapted? CAMMESA for each week. This price has a capShort run and long run marginal costs differ and equal to PPAD.this methodology would result in losses for thefirms with the less efficient equipment. This is Figure 2.2why generators are also paid for their capacity andfor their availability to the whole system. This Marginal Cost andcapacity payment can also be viewed as a private Inverse Demandinducement to decrease the probability of failure inthe whole system. For instance, if demand is very D3high like D3 in Figure 2.2, the only way to match \ \

this level of demand would be to accept a very -K---.-------------..'.high spot price P3 equal to MC3 . This payment

corresponds to "fixed cost" such as capital andsalaries, in contrast to the variable cost determined DEDby raw fuel costs. This fixed cost approximate the MC2cost of maintaining capacity to meet demand. It - - .-.-.---.-...................signals the need for extra capacity to generators. MCI

2.44 This signal is achieved by making a portionof the capacity payment increase with the risk ofunserved demand. When in any given week, the KI K2 generationLoss of Load Probability (riesgo de falla) exceedsa given value, generators get payments related tothe cost of unserved energy (Costo de energia no 2.46 Seasonal prices. Yet another common

sumincstrada) and fixed at 1500 $/Mwh. The problem the reform was able to address is theeconomic rationale is that as demand is very volatility in the retail price for electricity and theinelastic in the short-term, the system will be resulting speculative incentives. The spot price isunable to supply the whole demand. If the price calculated on an hourly basis. It can vary sharplycould not be increased beyond MC2 , an excess at times of peak demand. These demanddemand equal to ED in Figure 1 would remain, conditions in turn can vary with the business cyclesThis is why the capacity reward is so important. It and weather conditions. It can also vary on theprovides an incentive to increase capacity up to supply side due to fluctuations in raw energy

K2. prices and availability and the condition of themarginal generation units.

2.45 In practice, Argentina's generators get two

more types of capacity payments. For the 20 It is not clear that this reflects efficient pricing. In thecapacity made available (potencia puesta a UK, for instance, there is no capacity payment of thisdisposici6n), they receive a payment for each hour kind. This capacity payment introduces regulatedout of valley. The price paid per MW (precio de prices and the outcome is thus that the ES rather than

the market decides on investment needs.

21

the MEM operator about the volume it will have2.47 To reduce this volatility, CAMMESA also to dispatch, which facilitates the determination ofuses prices called "seasonal reference prices". the equilibrium price on the spot market.These prices, revised each three months, are the Contracts represent almost one third of the totalones charged to distributors and that are used demand for energy on the MEM.when computing retail sale tariffs. When theseasonal price is higher than the spot price, the 2.50 Like contracts between generators anddifference is credited to the distributor; it is distributors, contracts between large users anddebited in the opposite case. The outcome is that generators stabilize the exchanges in the short-runthe retail prices for electricity are less subject to and facilitate the planning of investments in theextreme variations. They reflect the changing long-run. The large number of this kind ofconditions in the bulk market only on a seasonal contract shows that the generation market behavesbasis. rather well. It could work even better if contracts

themselves could be exchanged on a market.2.48 Contracts to fix prices. However creativethe reformers were, they have not been able to 2.51 Transmission pricing. The general rule isreduce significantly the high volatility of the spot the adjustment to price increases with somemarket. Prices vary so rapidly that any short-term correction for improvements in productivity. Thisprovision is very difficult to perform. The solution is also known as the RPI minus x rule where x, theto reduce the impacts of this volatility in "productivity gains" variable, was initially set atArgentina's model has been to allow operators to zero in Argentina. However, the importance ofcommit through bilateral agreements. This creates large users in Argentina's regulatory frameworka futures market in the MEM where generators, shows that for any generator and distributor, thedistributors and large users can conclude contracts real key to transmission regulation is open accessfor energy and capacity at specified prices and to the network. Because Transener, the mainquantities. carrier, enjoys exclusive rights for the provision of

high voltage transmission (5OOkV) on the existing2.49 But there are many other advantages to network, it has to be regulated. Its regulation isthese contracts. For distributors needing to meet based on prices and quality standards.all demand in their concession area, contracts are a Distributors--and generators--are protected fromgood way to ensure they will not have to pay abuses through a very well designed and effectiveheavy penalties for defaulting in their obligation to system of penalties for non-compliance with itsserve or at least that they share the risk of established obligations, in addition to thedefaulting. Moreover, they can protect new prohibition to buy and sell electricity mentionedoperators of distribution companies by reducing earlier.the risks of having to deal with unreliablegenerators. Indeed, generators who have older or 2.52 The objectives of Transener's priceless reliable equipment do not like these contracts regulation are: (i) transmitting accurate signals ofsince they expose them to huge penalties if they costs from producers and carriers to distributorsfail to supply the contracted amount to the buyer. and large users and (ii) giving the operator goodIn sum, these purely financial contracts introduce incentives to develop the transmissionan element of stabilization in the wholesale market infrastructure. With respect to the first purpose,and introduces an instrument to hedge against prices are computed on the basis of energy andvolatile spot prices. Finally, another important capacity marginal costs. The short-run marginaladvantage is that contracts bring information to costs in the transmission of electricity are mainly

22

the physical losses of the line (proportional to the . An Energy charge is a third fixed chargesquare of the power transported) and the reflecting the difference between the value of"congestion" costs when the line cannot absorb all the energy received at a receiving node and thethe power injected by generators.21 value of energy at a sending node. The

fixation of this charge for a five-year period by2.53 The capacity costs are very hard to CAMMESA (with the approval of ENRE) iscompute because of the uncertainty regarding the to reduce the volatility of this source ofdates and length of "congestion" periods. To revenue. Its potential volatility stems from itsrelate positively the transmission-firm profits to dependence on the market value of electricity.the power losses on the lines is clearly not a verygood incentive for the infrastructure operator to 2.55 The connection and capacity charges areimprove the quality of its equipment, especially fixed for the first five years. Then they can bewhen it is a monopoly--i.e., the larger the losses, reduced yearly by ENRE to improve efficiency:the higher the profit and the lower the incentive to this decrease cannot be larger than 1% per yearinvest more. This is why Argentina's regulation with a maximum cumulative value of 5% in eachvery appropriately imposes quality objectives on 10 year management period. For the initial period,the transmission company. When incidents occur, energy transmission costs were fixed at US$55it pays penalties described below. On the other million for each of the five first years. The samehand, since short-run marginal cost for the procedure will be used for the next five-yeartransmission of energy does not include the periods. The energy charge is now fixed at $55equipment maintenance costs, Transener is also million per year for the first five years, payable incompensated for its transmission and connection equal monthly installments. Adjustment will becapacities. made as needed if a deficit or surplus arises.

When justified, transmission companies can apply2.54 In practice, these concerns have led to to ENRE for tariff adjustment. Similarly users canthree types of transmission charges, all of which complain to ENRE if tariff are considered too highare regulated by ENRE: and public hearings are organized which means

* A (Connection charge covers the costs of that many of the provincial private operators aremaintaining the equipment needed to connect likely to have regular interactions with ENRE.the user; it is a fixed charge (differentiated 2.56 Penalties, as mentioned earlier are a keyaccording to the voltage of the line connected, Pedientione egulary modelfrom $8/unit/hour for 132kV to $10/unit/hour ingredient of the success of the regulatory modelfor 500kV) and is levied by CAMMESA on all to make sure that the transmission companies dothe MEM members. not trade off quality for prices. Penalties are

applied for non-availability of either capacity* A Cal)acily chairge is also a fixed charge which and/or connections and apply to all unforeseen

distributes among all users the total costs of outages (planned outages have to be cleared bythe operation and maintenance of the existing CAMMESA and are subject to a discountedtransmission equipment. penalty). The level of penalties vary according to

outage duration, frequency, type of line and added_______ _______________ costs imposed on the system. The revenue from

1The cconoric concept of congestion refects the fact these penalties are rediscounted to users but arethat the transmission capacity installed is not always limited to 10% of the company's revenue over thesufficient to bear the loads that the market will bear. previous 12 months. If they are cumulatively moreThis is leads to power outages in which no one isserved by the transmission company.

23

than 15%, the operation license may be revoked by three months) and on electricity bought underthe ES. long-term contract with Central Puerto and

Central Costanera (adjusted once a year) (see2.57 Distribution tariffs. The distribution Box 2.5 on how the seasonal adjustment istariffs formulas set by ENRE where it has handled);jurisdiction were copied in all provincial a loss charge which corresponds to technicaldistribution privatizations so far. They include a and non-technical losses equivalent to aboutfull pass-through of MEM seasonal prices allowing 11% of the distributor's electricity purchases;for technical losses (11%) and a distribution costdefined in the contract. * the connection and fransmission costs;

* the cost of capacity in the MEM;M and2.58 The prices fixed before the monopoly wereconcessioned are used as price caps. This means * a distribution charge which is a fixed chargethat the price regulation system belongs to the for small users or a charge per unit offamily of the so-called RPI-x+y pricing formula maximum demand for larger users.which means that prices are allowed to increase ata rate equal to the Retail Price Index, less some Box 2.5 Dealing with the Seasonality of Pricesproductivity gain x, plus the increase y in the price The tariff reflects the seasonal price on the MEM.of some inputs on which the firm has no control This price, in turn, reflects, projections of supply and(essentially the costs of energy in the wholesale maximum demand by distributors and large users (for themarket). For set values of x and y, the prices are next two semesters) and information on generation

six months on the capacity, hydrology and transmnission restrictions. Theseautomatically adjusted every six months on the computations are done by CAMNESA. There is a winter-basis of the U.S. CPI and PPI (33% CPI, 67% PPI spring period (with high volatility because of rains andfor finished products). snows) and the summer-autumn period. The seasonal

charge ends up being a monthly charge fixed for three2.59 This pricing promotes cost-reducing months. It is based on the capacity demanded and theactivities but it can result in a decline in quality. corresponding costs.To prevent such a decrease, the concession The seasonal charge is designed to correspond tocontract organizes a complex system of penalties the spot price paid to generators in normal conditions. Butincurred by the distributor when it defaults. The since spot prices vary from the seasonal price, astabilization fund was established by ES to avoid over- orother problem with this type of regulation is that it underpayments by distributors to the MEM. The fund ispromotes a strategic behavior of less-than- used to pay generators when there is a deficit and to holdexpected effort when the managers know that the any surplus paid by distributors. It does not cover defaultspresent cost will be used to determinate the value by distributors. Such defaults are prorated to all generatorsof x for the next period. So the announcement of in proportion to their sales to the spot market. A failure to

the revision dates is a highly sensitive problem. pay for more than 30 days can lead to the suspension of thethe revision dates is a highly sensitive problem. distributor's right to make further purchases on the MEM.The productivity factor x is to be revised everyfive years, except for the first period of ten years. 2.61 The loss charge corresponds to aFor this initial period, x is fixed equal to zero. recognized allowance for technical losses. The

11% allowed may not cover actual levels of losses2.60 In practice, prices applied by distribution but reflects a reasonable benchmark. It is intendedconcessionaires in the RMBA are based on a few t

specific z~~~'Mnents. ~to give an incentive to distributors to achieve alevel of losses acceptable to consumers and

an energy charge which is based on seasonal regulators.electricity costs on the MEM (adjusted every

24

2.62 There is some degree of differentiation in Box 2.6 A Comparison of Tariffs across Provinces

the tariff. For instance, the needs of industrial A comparison of tariff levels across provincesusers are far more complicated than the needs of suggests that privatization may have been in the interest ofhouseholds. Consequently their valuation of many commercial and industrial consumers. Althoughelectricity is very heterogeneous and the there is a wide variation across consumer types, tariffs aredistributor provides them with a wider variety of in general lower in the areas served by private distributionoptions. So, the first market segmentation is companies than by public utilities. This is the conclusion

options. So, the first market segmentation derived from the analysis of a sample of eight distributionbetween small (<10kW), medium (between 10 and companies in August 1993 by H. Quifiones (published in50kW) and large users (>50kW). For large users, Novedades Econ6micas, December 1993, number 155,prices vary according the voltage of distribution: pp. 14-19). Comparing C6rdoba (EPEC), Greater Buenoslow, medium or high voltage. Only large users Aires (Edenor), Santa Fe (EPESF), San Juan (SESSE),have an hourly option. Like any other consumer, Mendoza (EMSE), Rio Negro (ERSE), Neuqudn (EPEN)

theypay tw-par taiff,butboththecapaity and Ushuaia, he found that the average commercial andthey pay a two-part tariff, but both the capacity service tariffs in public utilities were between 28% andpart and the consumption part vary according to 278% higher than for Edenor. For industrial users with athe peak or off-peak time period. capacity of less than 30 Kw, the average was between 84%

and 180% higher. For larger industrial users, the gap wasbetween 67 and 165%. Cordoba and Santa Fe were almost

2.63 The distributors have less flexibility on the systematically the most expensive in the sample. Andistribution charge. It is fixed by the regulation interesting additional finding is that, at least forbased on expected demand, and expansion and commercial users, there was no clear pattern in themaintenance costs. It can revised every five years direction of the variable cost component of the tariff.after the initial ten year period (although there is While it was decreasing in Bs.As., C6rdoba and San Juan,afte th iniialten earperod (lthugh her is it was constant in Mendoza and increasing in the othersome room for negotiation after two years if all cases.parties involved can benefit from a change). Itbasically includes the cost of distribution and a setrate of return for distributors. The update of thecharges reflects variations in the value of the U.S. 2.65 Since this pricing system is mostly a fixedconsumer price and wholesale price indexes. Theoutcome of this new methodology seems to have pnce one, distributors have an incentive tobeen favorable to essentially all consumer groups decrease costs and it could be at the expense ofas suggested by Box 2.6. quality. Also, distributors could be tempted to

differentiate their quality to induce the large users

2.64 Payment for service obligation. The not to contract directly with generators. To limitobigan for distfibutors to grant access to their this strategic behavior, the quality level of the

network to all those interested (essentially large transimssoon servtce ns controlled by ENREsusers) is a key ingredient of the system, but it Distributors that do not comply with the standardsinvolves a cost that has to be covered. This is why in terms of tension quality and number of cuts aredistributors are allowed to charge a toll tariff for penalized and receive reduced tolls.the use of their network. The charge essentially 2.66 Provincial regulators can preventcorresponds to the connection and transmission distribution utilities from cross-subsidy practicescomponent of the standard tariff. It also includes butin u ti the provincia f pructurethe loss charge. For provincial distributors, this by makig sure that the provincial tarcff structuretoll is fixed by the ES (through Resolution 159). part of the concession agreement iS consistent with

25

the structure proposed in the Resolution.22 As 2.69 As long as competition through directlong as the ES is able to ensure that this is done contracts between users and generators is viewedcorrectly there is no issue, but if in some provinces as beneficial, these tax barriers between thethe proposed tariff are too low as compared to the provinces should be source of concern.23 Withoutthe real costs, it may reduce the value of a these barriers, the user looking for a directdistribution company for a private investor. contract introduces additional efficiency in theAlternatively, if the tariff structure is much higher whole system through the competitive downwardthan the marginal cost, it can generate too high a pressure put on the profits of suppliers. As longrent that may be difficult to reduce once the as there is a perfect pass- through of wholesaleconcessionaires have gotten used to it. In Entre purchase prices to users, the distributors' businessRios for instance, after some fighting, a should not be affected by losing to a third party.cooperative negotiated a toll with the distributioncompany for $14,000/month. The value proposed 2.70 The distributors in fact do not have to loseby Resolution 159 was more to the liking of the large-users as customers. They can keep them bydistribution company: $180,000! giving a discount equivalent to the saving the large

user would make by contracting directly with2.67 One solution to this problem would be to generators. Other users may however not bereplace the value proposed by the Resolution (and indifferent to this approach. The distributor maycomputed on the basis of average provincial be tempted to avoid the cut in profits resultingconditions) by a methodology which would allow from a discount by increasing prices for the captivea better acknowledgment of the specific conditions consumers. This means that the regulatoryof each province. This is one of the conclusions authority has to check that captive users are notalso adopted by ENRE through its resolution 195 paying for this type of undue opportunity cost andof October 1995. being charged a disproportionate share of the

2.68 A riscal role. Distributors have also a common fixed costs. Within the boundaries2.68cAnt fiscal role. Distributors thve pals a determined by the avoidance of cross-significant fiscal role. Not only do they pay subsidization, the need to set some pricesnational taxes, but they also collect national, sbiiain h ne ostsm rprovincial and municipal taxes and these can be aggressively low in order to retain the business ofprovinial hand municout Tierra del Fuego, an large users means that other prices should bequite heavy. Ruling otTerdlFug,a permitted, within limits to be analyzed, to take upoutliner with hardly any tax, total tax rates for permitteffithientl lmsetorb anqal te yuresidential consumers vary from 18% in Santa the slack of efficiently securig adequate revenuesCruz or Misiones to over 40% in Entre Rios andLa Rioja. For most commercial and industrial 271 Financing the NationalnElectricity Fund.users, rates vary from just over 20% (in Santa A final role of the distribution company is that theyCruz and Misiones), to over 60% in La Plata are a collection agency for the National Electricity(Edelap) and Buenos Aires (Edenor). This leads Fund. They pay and collect mandatoryto inter-provincial tax differentials and ends up contributions to the National Electricity Fundcreating significant distortions in the competition Distribution companies and large-users payprocess.

-_____________________________ 23 The provincial taxes induce users in the province of22 The tariffs set by Resolution 159 actually imply large Bs.As. not to contract with a generator from an other

cross subsidies because they impose a single average province or on the contrary are a strong inducement totariff to large heterogeneous provinces. This may be bypass the local distributor through direct contracts (asunsustainable for many of them as below average EDENOR asserts) according to the buyer identity:customers are likely to opt to bypass physically the (consumer or distributor?).distributors.

26

$0.0024/kWh purchased from the MEM. Therevenue from this fund is divided as follows: 60% Investment and Tariffsto subsidize electricity rates for specific types ofregional consumers and 40% go to fund 2.74 While the reform introduced in Argentina'sinvestments in electricity in the underdeveloped electricity sector is impressive by any standard, asregions of the country (although all provinces are in any reform, there are some loose ends. Thein theory eligible). specific linkage between tariff design and the

incentive to invest may be one of these.2.72 Penalties. Distributors, just liketransmission companies, are accountable for the 2.75 Electricity is the only sector in whichquality of their services (product, technical and Argentina's reforms have introduced an incentive-commercial services). Their regulation includes a based type of regulation (composed of price caps,penalty system intended to make sure that minimum quality standards and clear serviceminimum quality standards are respected and that obligation) but in which the concessionaire hasthe investment needed to ensure the sustainability complete freedom to decide how much to invest.of the service will take place. Specific penalties In water or transport, specific investment targetsare attached to each type of non-compliance (e.g., have generally been set.tension at feeding points, frequency and durationof interruptions, billing mistakes, time required to 2.76 The specific linkage between tariff designrespond to connection and billing requests). If the and incentives to invest where it matters istotal amount of penalties is larger than 20% of net probably the main area where some fine-tuning issales in a given year, the concessions can be needed at the national level. This was a recurringrevoked. The ENRE is responsible for this theme in conversations held during variousmonitoring and for ensuring that penalties go back missions with all the relevant actors at the nationalto customers as rebates on their electricity bill. level. The linkages between tariffs and investment

are worth discussing in the context of provincial2.73 A generic pricing issue. The regulation needs because changes in these areas will affectof transmission and distribution, the two remaining the overall perception of private investors in thenon-competitive segments of the sector, is very sector. For instance, the mechanisms of expansionsimilar in terms of pricing. Basically, in both in transmission have direct implications forsegments, the tariffs are set in US$ and in both provincial distribution companies.areas, the regulation is a version of the standardprice capping formula (i.e., RPI minus x). In other 2.77 Investments in generation. Under thewords, the regulation sets a maximum price with current organization, generation investments aretotal pass-through of the costs of energy (Y) and decided independently by firms. This implies thatwith an indexation to the U.S. price index. The the type and size of new equipment depends onmain unresolved problem is how to recalculate x. private profit forecasts. How does the currentThe issue will already arise in two years, so it is regulatory environment ensure that this investmentimportant for the regulatory entity to start working is accurate when the generation market is an openon a methodology. If the British experience is of competitive market separated from theany relevance, this can be quite a challenging task downstream firm?likely to be contested by many of the privateoperators. It may in fact be reasonable to consider 2.78 In a competitive market, investmentthat this task be designed with the needs of the decisions are motivated by price signals. Theseprovincial distribution companies in mind, signals are sufficient to inform managers on the

27

users' willingness to pay and users on scarcity. For coordination between all the agents. It could bea firm, a price higher than the long-run marginal done through a joint venture if one accepts thatcost is the signal that users are ready to pay more individual firms pool costs, profits and risks. Anand/or to consume more, so that the firm should alternative would be to perform this internalizationincrease its capacity. From a social point of view, through the transmission operator since it is thethis investment decision has to be motivated by a essential link of the whole system. But this wouldpositive discrepancy between short-run and long- necessitate to change the ownership of the carrierrun marginal cost.2 4 So, for the private decision to and to transform it into a "cooperative" firm with acoincide with the optimal social decision, two representation of all users like in CAMMESA. Itconditions are needed: (i) the price has to reflect would be less attractive in the Argentina contextexactly short run marginal costs, and (ii) the firm but could provide a reasonable starting point inhas to know the long-run marginal cost. any other country considering a major reform of

its electricity sector. A simpler and possibly more2.79 Short-run marginal costs? It is not clear effective solution would be to revise the rules forthat prices fulfill these two conditions in capacity payments to speed up a decision processArgentina's MEM. First marginal cost pricing is that is now starting to be dragging in Argentinanot quite accurate since it essentially reflects the and may soon become a source of concern forcost of fuel (with some room for adjustment in potential investors considering bids for someother costs through Resolution 105/95) and not provincial distribution companies.the whole marginal operating cost as it should.Second, even if each generator does possess 2.82 Investments in transmission. Theinformation about the long-run marginal cost of its development of the transmission system wasown equipment (which is not sure), it cannot know mainly determined by the high concentration ofthe long-run marginal cost for the entire system consumption around Buenos Aires and thethat is a LRMC computed by taking into account remoteness of energy sources. The location ofthe whole demand load curve and the whole sector hydraulic resources and gas deposits far fromequipment availability. Buenos Aires in thinly populated regions required

the construction of very long high voltage lines. In2.80 Long-run marginal costs? Long-run the south, three 500 kV lines carry electricity fromsignals in the Argentine market are intended to be the region of Comahue to the Buenos Aires area.given through capacity payments to generators. If Two of these lines are strictly parallel and verythe value paid does not accurately reflect LRMC, close physically. The third one is never more thana distortion is introduced. In Argentina (for now), 200 km from the two others. This particularitynew investment in generation may be biased increases the risks of serious meteorologicaltowards base load generation with insufficient incidents simultaneously affecting the three linespeaking plants. Generators have a strong incentive and is just one of the reasons why the constructionto underinvest in units requiring huge sunk costs. of fourth line may be the only serious issue facingThis bias has also been observed in the United the sector in the short- to medium-run.Kingdom.

2.83 For now, the high-voltage network2.81 This bias could be corrected by its concessionaire (Transener) is not responsible forinternalization. This in turn implies some kind of the decision to build and to finance new lines.

This can be justified first because the natural24 The gap is financed by capacity payments. It may not be monopoly in charge of the network would have an

fine because it is calculated on technical criteria rather exorbitant power on all the other agents upstreamthan economic.

28

and downstream. The other reason is that the simple conceptual solution. Within certainoperator earns a fixed remuneration (for technical limits, a transmission line is a publicconnection, transmission capacity and carried good, that is an equipment that can be used byenergy) so that there is no distortion in the spot several firms without impeding the use by others.prices of electricity or in the prices fixed by This is why its financing is difficult. It entails "freecontracts. But with this cost plus pricing, the riding" incentives--i.e., not every user wants tooperator has few direct incentives to invest; the pay its fair share.indirect ones are penalties to be paid when it doesnot meet the quality of service established by the 2.87 Because a private user of the line will notconcession contract. internalize all the potential gains of the investment

to get an efficient level of investment, the2.84 This issue is, however, of utmost allocation of property rights must be made by animportance to the provinces. Because the entity with some responsibility for social concerns.reliability of the distribution service depends on the Also this entity needs to have some mandatoryquality of the SADI, the concession contracts need power to prevent free riding.25to spell out clearly if the pass-through ofexpansion or improvements costs to final users will 2.88 A realistic solution could be to give thatbe authorized or not. responsibility to ENRE but make sure that the

decision process is based on public hearings and2.85 Who makes investment decisions? In the that the Secretary of Energy is viewed as thecurrent system, decisions to develop the arbitrator in case of conflict. To solve free ridingtransmnission infrastructure are to be taken by the problems in the investment process, this entitypotential beneficiaries of the growth so that the should have the power to enforce exclusion ofcapacity component of transmission prices is "bad" agents (i.e., free riders) from the use of newcomputed on a stationary basis. The procedure to investments. To prevent abuses, an implicitexpand the high-voltage transmission system is the qualified majority rule would be necessary in thefollowing: investment decision process.

- the decision to expand is taken by the 2.89 Investment in distribution. This may bebeneficiaries; and

the most immediate concern of many provincial* building can be executed by Transener or by an governments, but it is much easier to solve than in

independent carrier that receives a technical the case of transmission. Decisions to invest arelicense. For a "small" expansion (less than left up to the concessionaire, in principle. Yet,US$2 m), the building is executed by because the concessionaire has a contractualTransener after an agreement with the obligation to service anyone who requests it at thebeneficiaries or the formal authorization by set tariff, the government has maintained someENRE to collect the costs. For large leverage on the decision process through theexpansions, the firm in charge of the design of tariff and penalties. The incentives are,development is chosen through a public however, well designed in the existing contracts asauction in which Transener is authorized to the tariff and the penalty system stimulate thecompete as any other firm, distributors to invest in an expansion of its

2.86 The construction of new transmission lines 25 In a vertically disintegrated electricity industry, besidesat an efficient level in a vertically disintegrated the Government, an alternative can be a commonindustry is a common problem with no clear and property of the firm in charge of the transmission

activity but this is not a realistic solution in Argentina.

29

network up to the point where the marginal cost of difficulties recently met in the discussion of theexpansion is equal to the marginal cost of fourth line. The current definition ofpenalties. This cost includes not only the actual beneficiaries responsible for the financing ofpenalties but also the revenue generated by the new line excludes large users; this is likelyadditional users of the system. Overall, the model to result in suboptimal investment indeserves replication in any contract considered by transmission as these users have quite a bit ofprovincial government considering the information they have no incentive to reveal toprivatization of their distribution companies. regulators, distributors or generators under the

current regulatory regime; while this problemSummary of the Main Recommendations is clearly one of the most complex in the field,

it may be under-analyzed in the Argentinean2.90 The chapter has shown how impressive context and deserves some more analyticalArgentina's reform of its electricity sector has work. In May 1996, the potential crisis wasbeen. This is particularly true in terms of the resolved by the decision by the government toimprovements of the short term efficiency goals of finance about $80 million of the $250 millionthe restructuring of the sector. The discussion of needed to finance the expansion. Theinvestments in the sector has, however, shown that generators will pay the rest. The expansionthere are some reasons for concern that deserve should be concluded in about 2 years. It willattention. The fine tuning needs identified in the help in the short run and should lead to lowerchapter, well known to the ES and the ENRE but wholesale prices but it does not yet provide amaybe not to provincial authorities, include: long term solution to the transmission pricing

problem.

Generation = A medium run issue is the need to get ready to

* A better definition of short run marginal costs adjust the x in the RPI - x formula used tocould include, not only fuels, but also a better regulate prices; this process can require someassessment of other components such as the time in consultation as seen from the Britishcost of the marginal generating unit (not just a experience.15% margin over fuel costs).

* Contracts are already very effective in reducing Distributionshort-term price volatility, they could be even * As more and more provinces privatize theirmore efficient if they could be traded on a distribution companies, the relationship"contracts" market. between the transmission concessionaire and

* The pricing of capacity should signal long-run the privatized concessionaires is likely tomarginal cost pricing but the current design of change and the tolls may have to be revised tothe capacity payment is biased against ensure consistency with marginal cost pricinginvestment in generation with high sunk costs by the privatized distribution companies;(as was also observed in the UK). * Unless a clear pricing mechanism with

complete pass-through of wholesale prices isTransmission adopted, the captured users may end up paying

.The major transmission issue (in fact the major the price of the losses of distribution profitsissue in the sector) centers around the linkages due to direct contracts between generators andbetween tariffs, property rights and investment large users; this may lead to difficult politicalneeds. Clearer rules are needed to avoid the problems for privatizing govemments as tariffs

30

post privatization may end up being higherthan before for a large share of the electorate.

* Just as in the case of transmission, theregulators need to get ready to recomputesoon the "x" in their regulatory formula; theten years wait for a revision of the "x:" may infact be unsustainable and a revision may haveto be considered sooner. In the short-run, theeasy way out for the provinces trying tointroduce more efficient pricing is to adhere tothe tariff principles established in the law (andnot necessarily apply the numbers establishedin Resolution 159).

Reeulation

* The contracts for distribution should remainthe main basis for regulation and shouldcontinue to be based on the successful modelimplemented so far by the ES as part of theirtechnical assistance strategy to the provinces.

* The provincial regulatory environment needsto be spelled out clearly, kept to a minimumand coordinated with the national regulator toprovide as much relevant information aspossible to potential investors.

31

Chapter 3: DELIVERING WATER AND SANITATION SERVICES IN THE PROVINCES

3.1 Just as in the electricity sector, the main coordination and regulation of the sector and itspurpose of the reform initiated by the National pricing policies.Government in the W&S sector with theconcession of W&S services to Aguas Argentinas Table 3.1 Impact of the W&S Reform in the(AA) in Greater Buenos Aires in May 1993 was to Greater Buenos Aires Areaimprove efficiency and to get the private sector to Changes between May 1993 (Ist monthfinance expansion needs. A major difference is of operation of Aguas Are tinas) andthat the impact of the national reform has been Dec. 94 Dec. 95

% increase in Production 25% 26%much more localized. The sector is indeed capacity (000M3/day)decentralized and the national reform has been Kmnofwaterpipes 200 550

restricted to the federal capital and the 13 districts rehabilitatedaround it where the National Government has Km of sewers drained 2,150 4,800jurisdiction. % decline in number of 81% 97%

clogged drains3.2 But this experience demonstrates that the Number of meters upgraded 68,200 128,500

W&S sector presents a particularly difficult and installed 48% 47%

challenge for the provinces. On the one hand it is Number of new residents 350,000 642,000

a natural monopoly for which the scope for direct with water connection , _

market competition is limited and hence for which Number of new residents 200,000 342,000

the regulatory environment is just as important as with sewer connections l

it is for electricity distribution. On the other hand, Source: Aguas Argentinas

it is a sector for which the value of the assets ismuch more difficult to assess since most of it is in 26

the form of underground pipes. This implies thatthe risks faced by the private sector are much 3.4 Under Argentina's constitution, thehigher than in the electricity distribution business responsibility for W&S rests with the provinces.

where the assets are easier to assess. In spite of this, until recently, the administration

3.3 These risks have already been considered and planning of the sector was highly centralizedworthwhile ,., private operators in some in the hands of Obras Sarntarias de la Naci6n

wrorthwhil byoprivates has suboperatoes any som (OSN). The first major reform of the sector tookprovices. Corrientes has sboncte manyaof place in 1980 when sanitation in Argentina was

oethe oW services inne t0 pormovance Snto p te effectively decentralized to the provinces. Theoperators since 1990. Formosa and Santa Fe .oine .nl manaie th oraiatoarecently completed their own privatization. The ge y g

structure set up by the National Government whovarious experiences so far provide enough followed a deconcentration strategy of itsevidence that the reforms are having many of the activities. In some instances, the provinces furtherexpected payoffs (see Table 3.1 for evidence on . . .the changes brought by AA in Buenos Aires). Butthey also highlight the issues that other provinceswill have to address as they try to follow the 26 A recent World Bank report provides an overview ofprivatization road. These center around the the sector organization: Argentina: Water Supply and

Sewerage Sector Note, LAlIN, July 27 1994.

32

The result of these changes is that there are now . The Secretariat of Natural Resources andabout 1500 operators in the W&S sector including Human Environment (SRNAH) which ismany cooperatives. In El Chaco, Chubut, La responsible for the environment since 1991.Pampa and San Luis for instance, most of the Its mandate is to coordinate environmentalservice is provided by cooperatives. issues among federal agencies.

The "Regulators" . The provincial Public Works Ministries andSecretariats.

3.5 Regulatory objectives. In addition to the * The provincial environmental Secretariatsefficiency goals of the reforms decided by the which have direct responsibility forNational Government and at least three provincial environmental protection, an area that isgovernments, the regulators have to address many mainly in the hands of the provinces.public concerns. These include environmental and

social~.. cocrs.hsmliliiyo ol a Regulatory agencies such as ETOSS (Entesocial.concerns. This multiplicity ofTripartito de Obras y Servicios Sanitarios) for

resulted in a multiplicity of agencies with some t 2 7

frm of reuatr resposblt. the Greater Buenos Aires area. In terms ofthe purely economic regulation, these areindependent agencies. ETOSS represents the3.6 The actors. There is no single agency itrsso h ainlGvrmn

comparable to the Energy Secretariat responsible inved as the Neriofah asset tfor a broad policy formulation and in charge of Muniplity an the Province of Bs.As)

'din .netie an fiacig Thr ar at Municipality and the Province of Bs.As.guiding Incentives and financing. There are at ETOSS' objectives are to regulate, enforceleast a dozen important public actors in the sector. g -The main ones are: regulatory decisions and fine when necessary,

owatch over quality and continuity of water and* The Puiblic Works Secretariat (PWS), sanitation service delivery, to protect health,

responsible since 1994 for the preparation and hydro resources and the environment and toimplementation of a national water resources oversee that the specifications of the contractpolicy and for the coordination of projects. in terms of the development plan are met. It is

* The Subsecretariat of Water Resources (SWR) currently financed by revenues from a user feewithin the PWS, which has the actual the concessionaire levies on all the consumersresponsibility for these tasks. The SWR is and is ruled by a Directorate whose members

are appointed by their administration for a six-also responsible for the promotionoer al year duration, with the option of one

p,ole ais the mainc renewable term. With a concession durationdialogue withtheWrldBanof 30 years, the risks of regulatory capture

. The Ente Nacional de Obras Hidricas de cannot be completely neglected--the longerSaneamiento (ENOHSA), an entity under the contract the larger this risk tends to be.discussion in Congress is expected to replacethe Consejo Federal de Agua Potable ySaneamiento (COFAPYS). Its main role willbe to mobilize and channel financial (mostlyinternational) resources to provincialsanitation companies to local communities in 27 In the ETOSS board, there are two members of each ofexchange for significant reforms in the sector. the following: National Executive Power, Provincia

de Buenos Aires, Municipalidad de la Ciudad deBuenos Aires.

33

3.7 Other national actors include the Interior contract. Just like the global contract, it grants aMinistry (involved in interprovincial rivers), the 30-year concession to a private operator withHealth Ministry (who sets norms for drinking specific coverage and investment objectives. Thewater quality and assesses the environmental main difference is that the contract does not coverimpact of infrastructure projects and manages a the whole area previously under the responsibilityregister of polluting sources), the Economic of the public utility. But in fact it covers the areaPlanning Secretariat and the Social Development that used to be covered by the public utility,Secretariat (who finance some projects or provide leaving to cooperatives the rest of the service.technical assistance to the provinces).

Box 3.1 The Timing of the Privatization Process in GreaterContracts as the Main Regulatory Tool Buenos Aires

The reform of the provision of water supply and3.8 Three main types of contracts have been sewerage services to Greater Buenos Aires started in June 1991bid in Argentina: and ended in May 1993 with the actual transfer of the company to

the winning consortium. It started with the work of an 11* global concession contracts; member Privatization Committee (including all government

agencies but also unions and Congress members). It chose toe partial concession contracts; and concession rather than to sell because of the desire to avoid any

D "management" type contracts. delay in the reform. The Commission also decided to limit thescope of the coverage of the contract (drainage, for instance wasleft to the municipalities as typically done in Argentina rather

3.9 The 30-year concession for the Great than included in the contract) and ruled against the possibility ofBuenos Aires awarded to Aguas Argentinas in creating two or more concession areas discarded for technical1993 is typical of the first type of contract. The reasons.

concessionaire assumes responsibilities for The preparation of the actual bidding documents and ofoperating, maintaining and investing in fixed the model concession contract was left to two specialized privateassets which remain the property of the public consulting firms. The first prepared the draft regulatorysector over a specific service area for a specific framework, the technical aspects (including performance, quality,

period. Over the concession period, the operator metering, rates targets, obligations and the prequalificationperiod. Over the concession pefiod, the operator criteria), the rehabilitation and expansion requirements and ais obligated to extend service coverage, guarantee review of all related laws and decrees relevant to interestedwater quality and develop sewage treatment. private parties. The other firm focused on the financialAccording to the concession contract, the (including the consequences for rate of various implementationconcessionaire has to assume all commercial risks. options) and marketing aspects. They also spelled out theBut there remain some renegotiation possibilities methodology to be used to award the contract. The call for bids

was sent in June 1992. The consulting firms assisted thethat allow redefiming the rules when some Committee in the evaluation of the bids of the five internationalunforceable event or some new information consortia who had prequalified (each had paid $30,000 for theoccurs. The details of the contract are bidding documents). The bid opening followed three monthssummarized in Table 1. la and 1. lb (Chapter 1) later and the award was decided in December 1992. It tookand Box 3.1 provides a brief overview of the another three months to sign the contract because of the timestages required to be able to conclude the required for the winner to present guarantees and insurances and

a decision to allow the winner to start operations in goodprivatization and Table 3.2 shows the main conditions rather than during the summer which is typically adifferences between the contracts of Corrientes, problem season in terms of public relations due to unsatisfiedTucuman and Santa Fe. demand.

3.10 The model adopted by Corrientes could beviewed as an example of a "partial" concession

34

Table 3.2 Main Provincial Water Contract.. ~~~~~~~~~~~~ ................ .................. ........................... ...........................

:::::::::::::::::: ......... . .contracts type concession concession with equity of 25% concession

required and 3 years experience inproviding services to at least Imillion people

duration 30 years 30 years 30 yearsbids 2 stages; stage I for technical 2 stages; stage I for technical and 2 stages; stage I for technical and

and financial pre-qualification financial pre-qualification and stage financial pre-qualification includingand stage 2, proposal on tariff (as 2, proposal on average tariff for requirement of a guarantee of% of current) successive years of concession (as % financing of credit institutions for

of current) USS 100 million to be disbursed inS yearsand stage 2, proposal on tariff percubic meter, debt service coverageof at least 1 and debt to revenueratio of 1.5.

investment $1 I million in first 5 years $20 million in first 5 years $100 million in first 5 yearsneeded $75 million in 30 years $312 million in 30 years $1,200 niillion in 30 yearscustomer base 500,000 inhabitants in 10 cities 700,000 inhabitants in 120 urban 1,750,000 (Santa Fe city and

for water and sewerage communities for water and 327,000 Rosario)people in 10 comimunities for

_ ~~~~~sewerageresponsibilities O&M as well as investments O&M as well as 100% water O&M with minimum target andof private coverage after 6th year and 100% objectives set in the contract and aoperator sewerage services after 11th year + maxinum time frame for each goal

emergency work in first 2 years (expansion, quality, pressure,______ ______ ______ ______ _____ _____ ______ ______ ______ _____ rehab., treatm ent)

The provincial governments maintain the 3.11 The third type of contract is a form ofresponsibility for service delivery in the more management contract. This has been a solutionisolated locations where full cost recovery is initially considered by Mendoza. It is not a pureunacceptable for social and political reasons. concession contract in the sense that it includes aSome have retained the Provincial Service for financing obligation imposed on the managementDrinkable Water and Rural Sanitation (SPAR) to company. But is not a pure management contractdirect international resources to rural water and either to the extent that the private operator has asanitation projects in areas excluded from the financial stake in the service company. Theconcession contract. The approach has generally proposal for Mendoza required that theworked so far for Corrientes. Since the private management company bring with it a capitaloperator took over in September 1991, water equivalent to 20% of the current level. But theservice coverage has increased from 59% to 71% idea has now been rejected by Mendoza where theand sewerage from 30% to 37%. Metering has government has decided to sell 51% of the sharesincreasing from basically nothing to 68,600 meters of the public utility.and the personnel has creased from 5.7 employeefor 1000 water connections to 2.9. The main 3.12 The bids were organized in March 1995problem has been to find financing for further and seven candidates showed interest. Theexpansion. management contract is for a duration of 20 years.

The provincial government will initially keep 70%

35

of the shares and 10% will be made available to obligations are not only spelled out in terms ofthe employees. Within a few years, 9% of the target but also in terms of processes. In the casegovernment share will be put up for bids. of unjustified and repeated cuts in service or ofAdditional requirements include a commitment by delays in investments (i.e., non respect of thethe private investor equivalent to $100 million (for PMES), the contract specifies a system ofa total investment plan of $420 million) and penalties.proven experience in the management of similarservices in regions with more than one million Box 3.2 "Goal-Based" vs. "Process"-Based Contractspopulation. Because the Bank mission did not get a copy of all

the concession contracts, a detailed analysis of the various3.13 The first two types of contracts are the designs was not possible. But interviews with some of themost common in all the ongoing privatization private companies who participated in all the bids provided

efforts. The main problem faced by provincial some useful insights on their perception of the maingoveminent in these efforts to get the private differences between the various types of contracts. One of

government m these efforts to get the prlvate these differences was that the concession contract forsector interested in these contracts stems from the Greater Buenos Aires was criticized for focusing too muchfact that the state of the "hardware" (i.e., the on the "processes" (i.e., multiple partial goals which can bepipes) is unknown. A good part of Argentina's used to "micromanage"), rather than focusing only onwater systems--including water treatment plants performance targets. The private operators also want toand large water mains--are more than 60 years old have some discretion in the disbursement strategy for their

investment commitments. The reassuring message fromand need replacement. Maintenance has generally the private operators was that the more recent contractsbeen poor and there are significant water losses have learned the lesson and were much more focused onthroughout the distribution systems. On average, targets than on processes.the unaccounted water in the urban water systemsis estimated at over 50%. 3.15 The remaining issue is that it is not clear

how much the concessionaire will have to3.14 The contractual obligations are designed rehabilitate the system and who is going to pay forto address these problems and include maintaining it. Aguas Argentinas in Buenos Aires is findingand developing the existing system for instance. out about it the hard way. It may haveThey include not only general tariff principles underestimated the rehabilitation needs. To avoid(discussed below) but also service quality and this problem, Santa Fe got an independentinvestment requirements with a specific timetable assessment of the state of its assets beforeand competition-oriented procurement rules (i.e., organizing the solicitation. This provided ainvestments have to be bid out). Failures to common set of information to all bidders (whocomply, as in the case of electricity lead to actually paid for it in the process).penalties. The way the bidders are going to dealwith these requirements is described in their 3.16 What Santa Fe found was a way oftechnical offer in which candidates show a reducing total risks, but this may not always berehabilitation and investment plan (PMES).2 8 The enough. An alternative solution may be toPMES generally describe their objectives in terms consider different levels of risk sharing betweenof connections (to drinking-water supply systems the government and the concessionaire. This canand sewers system), in terms of effluents be done as part of a negotiation and the highertreatment and of networks renovation, although as share of the potential cost goes to the less riskexplained in Box 3.2, in some cases, the averse party--e.i., probably the licenser. As the

end of the concession approaches, this requiresthe auditing of the investment program to become

28 Plan de Mejorasy Expansion del Servicio.

36

tighter as neglect of the concessionaire's was a sealed bid auction based on technicalobligation to reduce service costs becomes an qualifications of the bidders and with the contractincreasingly attractive option if it is considering granted to the lowest price for service delivery.pulling out. Such a strategy would increase thegovernment's share of the potential costs of 3.19 More specifically in B.A., the candidatesrehabilitation. compete by proposing a discount coefficient to

apply to the "public tariffs" followed until recently3.17 Management contracts may have been by the government. This is equivalent to awardingunderestimated in the Argentina context. Because the bid to the operator willing to charge theof their fiscal situation, a tradition of controlling lowest average tariff, as was done in Santa Fe.of water prices, or maybe just because of the very The bidding process was effective in reducinglimited information on the state of the companies' costs although not by as much as everyone thoughassets and hence on the actual rehabilitation needs, since costs went up faster than any one hadsome provinces will represent too high a level of anticipated. It turns out that, although on averagerisk for many potential concessionaires, unless users are better off, both the government and thethere is some type of guarantee imposing a limit bidders may have misjudged true costs (seeon the risk level they have to face.29 Few Box 3.3).provincial governments have the capacity ofgranting this guarantee which is why management Box 3.3 What Adjustments in the Terms a Contractcontracts can be such an attractive solution. A Revealsshort-term contract (e.g., two to three years) On June 1994, the National Government and the

would give a potential concessionaire an concessionaire (AA) added some service requirements toopportunity to gather and publish as much the terms of the contract signed 23 months earlier:information as necessary while operating the investments has to be made faster than initially thought.sector as a private utility. At the end of the term, In doing so, both parties revealed that they hadthe province could assess how much it needs to underestimated the degree of urgency of the investmentcommit to attract a private operators into a program. Adjustmnents needs or mistakes happen and

this is why rules are necessary for the renegotiation ofcontracts. One such rule was included in A.A.'s contract.

who wins the management contract could have a For this new definition of the conditions of the contractright of first refusal to ensure that it has no and as a result of cost increases (which had accumulatedincentive to hide the information on the assets to levels that triggered a contract clause that allowed avalue collected during its management contract tariff increase), A.A. obtained a 13.5% tariff increase.tenure. This has lead ETOSS to be more watchful of A.A. supply

and service contracts to make sure that no collusionwould occur between the concessionaire and its main

How Competition Works suppliers.

3.18 The main instrument to improve incentivesin the sector is competition. This competition is 3.20 Introducing yardstick competition. Aso far only achieved through the bidding process. better solution, so far ignored in Argentina'sAgain, the national experience is representative: it context, is to rely on some form of yardstick

competition. The provincial basis of the sector29 See the background papers by Humplick and Nasser provides an opportunity for the regulators to use

(1995) for an assessment of the various sources of comparative information on firms in setting pricessector specific risks in each province. See also Freixas and quality standards. This would require aand Garcia-Fontes (1995) for the financial engineering coordinated regulatory capacity to ensure theoptions that could be used to generate guarantees.

37

comparability of data. This is done very risks onto its shoulders. As mentioned earlier,effectively in the United Kingdom where water these risks are high as very little is known on theprices are regulated according to yardstick value of the assets and on the problems likely tocompetition mechanisms.30 It also requires a be met in meeting quality and quantity obligationsdetailed analysis of the factors that may explain of the contract.differences in operating costs and hence in tariffs(see Box 3.4). Box 3.4 What Yardstick Competition Entails?

It starts with a comparative efficiency review.3.21 A potential additional source of This review takes into account the influence of variouscompetition. A final way of introducing factors in the companies' operating environment thatcompetition in the sector is to separate the sector might explain differing operating costs. These factors arevertically into separate water and sewerage then combined in an explanatory factor index. A linearvusinessertically organize separate bids. Thiscan relationship is assumed between this factor and the unitbusinesses and organize separate bids. This can costs. This then allows a better comparison of thebe easy when the two services were separated to efficiency of individual companies. In England, thebegin with as was the case in C6rdoba. Similarly, companies were placed into four efficiency bands and thesewerage services can be divided into the banding determined the assumed possible reduction incollection business and the treatment and disposal operating costs. Each one was supposed to be at least

business. Ts .iintroduces capable of reducing costs by 1%. While there arebusiness. This disintegration problems with this procedure, it is a clear improvement

competition and also reduces many opportunities over the current practice in Argentina. Additional detailsfor cross subsidies in a sector prone to it. There can be found in Armstrong, M. S. Cowan and J. Vickersare however many reasons why it may often make (1994), Regulatory Reform: Economic Analysis andsense to maintain these activities vertically British Experience, MIT press, Cambridge.integrated. The costs of metering and billing forinstance would increase in a disintegratedindustry. But more importantly, many economies 3.23 The main requirement built in the contractof scope would be lost (in the pipes related is that the tariff must:works) and it would be harder to force the . promote a rational and efficient use of theinternalization of the pollution costs imposed by services and resources made available;the sewerage business on the water business.

equilibrate demand and supply;Pricing . meet the sanitation and social goals spelled out

in the contract;3.22 The general pricing principles followed inBs.As. are typical of those mentioned in contracts * reflect the economic cost of service deliveryin Santa Fe or in Corrientes for instance. They including the profit margin of theessentially correspond to contracts trying to get concessionaire and the costs imposed bythe private sector to accept some of the risk approved expansion plan;sharing by allowing them to charge for all . allow cross-subsidies;3 1 andrecognized costs plus a profit margin. A pricecapping approach as used in electricity distribution * not discriminate between users within usersis generally believed to be more effective at categories.inducing efficiency but would be much lessattractive to a private operator as it shifts all the

31 Cross-subsidies are not allowed in the contracts signed30 See Armstrong and alii, 1994, p. 341. in Formosa and Misiones.

38

3.24 The regulatory regime recognizes costs 3.27 Users without meters. For residentialand allows for an additional profit margin (cost users without meters, the bimonthly bill is givenplus regulation). In practice, the combination of by the basic tariff formula:the bidding process and these restrictions boilsdown to an endorsement of average cost pricing. TBB = K * TG * HThe concessionaire sets the tariff values within the where H is ani function of the surface ofrestrictions imposed above. It can also revise houses and non constructed lands. It is a proxytariffs in several circumstances. First, there are.. . . ~~~~~measure of water consumption during two"ordinary revisions" associated to changes In the mesrofwtrcnupindigto

months. For housing, it also depends on theservice Improvement plans integrated in the l ancconcession contract. No such ordinary revision locathe degree of luxury and implies a subsidy fromcan occur during the first five years. During the the m os luxuri residet e re(following five-year period, prices can only be ratio is 1.51 to 0.6). The relation with the age ofdecreased and during the subsequent periods, the s is non-ina. The tG ariablrevisions occur only if the concessionaire has ,,st ruthe "generari and K visbtecomplied with the plan objectives. Second, there ct int on whi competitor ad t id forare "extraordinary revisions" when some specified concession wing Inm 194,tG was equalthe concession winning. In 1994, TG was equalcost-index increases (or decreases) by more than7% and/or if the administration or the regulator to 0t0279 $/m3 for water plus the same amountchanges some basic conditions on which the for the sewerage connection. These are theconcession contract was signed, for example amounts for residential users and non-constructedquality standards. With the authorization of the land For non-residential users, the general tariffsregulator, the concessionaire may also modify the were doubled. Since TBB is expressed in Pesos,tariff system insofar as it does not result in a K is a pure coefficient and TG is expressed invariation of the revenue per capita. This is how $/m3 , H has to be expressed in m3 .price regulation allows for some cross subsidies.

3.28 The outcome of this form of presumptive3.25 Tariff differentiation. The regime taxation is that the total bill of these users isrecognizes three types of buildings: residential, independent of their actual consumption and therenon-residential and land without construction. is some degree of progressiveness in the "tax"The first two categories are then divided into used to pay for the cost of the bill as the richer thebuildings with one or more consuming units. In consumer is presumed to be, the more he/shethe latter case the bill is divided between the pays. The problem is that it provides very littlevarious units. This leads to four types of tariffs information to the concessionaire on the specificstructures. In the Buenos Aires area, 85 % of sources of water losses (see Box 3.5).consumers are residents and 15% are industries.

3.29 Users with meters. For the users with a3.26 The way in which the tariff is charged also meter (and the concessionaire would like todepends on whether the consumption is measured maximize this number as seen in Box 3.5), theor not. For industrial users, meters are mandatory tariff is essentially a two-part tariff with a fixedwithin two years of the beginning of the contract. part to pay for the infrastructure (TBB/2 or halfFor the user without a connection, a fixed tariff the basic tariff) and a variable part proportional tobased on the area covered by the building and consumption C (in m3 ) and derived by multiplyingIO% of the land is collected. This is the bimonthly the quantity consumed by the unit cost ofbasic tariff (TBB).

39

providing drinking-water (including reprocessing).The bill is basically calculated as follows: Figure 3.1

Bm = TBB +(K * P * C)2m Slope K *P

In 1994, the value for P in the water bill was

equal to 0.33 $/m3. For the joint service ofwater and sewerage, P is doubled. Actually, for TB- - .

residential buildings, there is a "free

consumption" block of the first of 30 m3 over a TBE.two months period. 2 .C

Box 3.5 Why does the Concessionaire Want Meters 30 .H/24+30Everywhere?

For the Greater Buenos Aires area, the current 3.31 For a consumer with given housingestimation for water production is 580 liters per capitaper day. The corresponding consumption level is only charactefistics (i.e., with a given H), installng a325 liters per capita per day. This reflect the fact that meter is a wise decision only if consumption islosses and illegal connections represent about 44% of small. For a consumption larger than C, payingtotal production in Greater Buenos Aires. the fixed price is less expensive. So, in each H

With the generalized installation of meters,Aguas Argentinas and ETOSS hope that production will ,be drastically reduced to satisfy the current consumption. to ask for a meter.First because the losses coefficient will be reduced from44% to 25%; second because the connection and 3.32 In sum, if people can choose between thecorresponding billing will induce a decrease in fixed bill (without meter) and the two-part billconsumption. The anticipated decrease in production (with meters), it is equivalent to having a menu ofcould be from 580 to 333 liters per capita per day i.e ta pe

O , ., ~~~~~~~~~~tariffs performing a self selection between users.43%. In the longer run however, the expansion ofservice should lead to an increase in total production. The drawback is that this selection process is

biased in the wrong direction since only the small3.30 The marginal consumer. A comparison consumers have an incentive to install a meter.of the bill for a residential consumer with and For water, an average price that is lower for thewithout meters shows that it is the same when his large user than for the small user cannot beconsumption verifies the equality Bm=TBB or justified on technological grounds, contrary to

2 + [K * P * {C - 30)]= TBB ~what exists for electricity (see Box 3.6).2B +K *P (- 5-30) = TBB

3.33 Another problem with the current two partConsequently, the indifferent customer is the one tariff is that there is no evidence that the total

Hwho consumes approximately C - + 30, as connection costs add up to the revenue from

24 TBB. Since the costs of installing and maintainingshown in Figure 3. 1. meters are well known, it would be reasonable to

expect that the fixed part of the two-part tariffreflects these costs to avoid a further source ofdistortion in the choice between metering or notfor residential users and to avoid potential cross-subsidies for metered industrial users.

40

Investments and Tariffs Box 3.6 What's Wrong with the Average Price ofWater in Argentina's Provinces?

3.34 A significant problem with the tariff One of the advantages of metering is that itformula not mentioned so far is that it does not allows a smoother discrimination among users andprovide much incentive to invest in the expansion makes it easier to send signals consistent with the socialof the sewerage system. This comes from the fact needs of society. These needs include a concern for the

poor and a concern for the environment. The concern forthat the current design of the metered tariff the poor would lead to expect a decreasing average pricegenerates a complex system of cross-subsidies for a limited level of consumption. The concern for thebetween water services and sewerage services. environment would lead to expect an increasing averageWhatever is paid for water is what will be charged price for water when there is a scarcity value attached tofor the sanitation component of the contractual this water. It turns out that the current two-part tariff forobligations of the concessionaire. For now, the metered users does a good job at meeting these.

evidence is that the system penalizes the sewerage For C <30m3 , the average water tariff iscomponent of the tariff As documented in a obviously decreasing.recent Bank report, household sewerage is When C> 30 m3, the average price is given by:probably the most important source ofenvironmental pollution in Argentina.3 2 This may TBB-K 30be the best indicator of a problem in the tariff 2 +(K *P).

design. CIt is an increasing function of C if:

3.35 The key policy issue here is how shouldthe regulatory framework be designed to TBBencourage the development of least-cost operation 2 -K*P. 30<0.and expansion of the sewerage network. There is

clear videnc of wilingnss to ay fortheseFrom the formula for EBB and the respectiveclear evidence of willingness to pay for these values of P and TG, this condition is satisfied if H< 709.services (see Box 3.7) so where do incentives fail? For most common households, this is likely to be

satisfied. Effectively, even if the H values are verysensitive to the type of building, most households arelikely to be below the threshold value.

Consider the current estimated consumption of325 1. For a four person family during two months, totalconsumption averages 78 m3. If the H estimation is nottoo divergent from this average consumption, the averageprice of water will be increasing for most users. Only forlarge users with very high H, can the average water pricebe decreasing.

32 The World Bank (1995), Argentina: ManagingEnvironmental Pollution: Issues and Options,May 24, in 2 volumes, Report No. 14070-AR. See inparticular Chapter II of the Volume II.

41

Box 3.7 Willingness to Pay for Sewerage costs from new customers and addressing more

According to an analysis conducted by COFAPYS carefully some social issues (see Box 3.8).3in Bs.As., C6rdoba, Santa Cruz and Santa Fe, thewillingness to pay for sewerage services is high. It Box 3.8 Dealing with Social Concernsaverages $35-55 per month per household. This is about7-12% of the income of a four person family at the poverty There are two main strategies to improve serviceline (about $102 per capita per month). Yet only 38% of delivery to low-income groups: low-cost services andthe population gets sewerage services (ranging from 10% subsidies. With regard to low-cost services, reform doesthe opuatin ges swerge srvies rangng rom10% not necessarily reduce the options available for locallyin Misiones to 73% in Tierra del Fuego). Cooperatives areincreasingly providing the service for a typical tariff appropriate choice of technology. Concession contractsranging from $3 to $27, depending on the type of service can allow the selection of appropriate technological optionsprovided. It remains well below the willingness to pay based on community participation and the increasedand has contributed significantly to reduce the coverage likelihood of bill collection if service costs reflect thegap in many provinces. More detailed studies are now respective communities willingness to pay.being conducted by ETOSS which may result in significant To subsidize low income consumers, explicitrevisions but for now this is the impression that was lump-sum subsidies for the basic consumption of low-conveyed to the mission. income consumers are preferable to tariff cross-subsidies.

If water companies can charge the subsidized part of the

3.36 The main incentive problem under the full cost of service provision to poor consumers directly tostandard two-part tariff formula used in Argentina the governent, their incentives to serve poor consumers

is that current water and sewerage charges arehigher than operational costs but do not cover thefull investment costs. This arises because Summngup theMai Recommendationscontracts sometimes allow that the share ofrevenues from existing customers that exceed 3.38 The chapter has shown that theoperation costs be used to cover the share of concessioning of water and sewerage services arewvater system expansion costs not covered by new likely to achieve some of their efficiency andcustomers but does not require any targeted level investment goals. To maximize the achievement aof sewerage coverage. The size of the cross few adjustments will have to be made in futuresubsidy from existing to new customers is in the concessions or in the renegotiation of existingorder of $480 million in present value terns contracts. The main aspects of the contract thataordingtof480 iBa illionk st tin esent value terms need revision or better consideration are in theaccording to Bank staff estimates. tariff design. But to be successful the reform will

3.37 In the case of Aguas Argentinas, the also have to consider whether an overall andproblem was solved through an adjustment in the major institutional reform is needed to clarify thedeal. 23 months after the signature of the sectoral regulatory and planning responsibilities.contract--to some extent allowed by one of the More specifically:

contract clauses. A 13.5% tariff increase wasgranted to AA in return for: (i) the advancementof service expansion; (ii) the provision of drinkingwater to some low income neighborhoods; and(iii) the replacement of water from some nitratepolluted wells. This provides the resources in theshort-run, but a more fundamental change in thetariff structure would be more desirable for allnew contracts to be considered in Argentina. This 33 The options for how to address the need of low incomewould involve recovering the long-run marginal neighborhoods are addressed in World Bank (1995),

Argentina: Managing Environmental Pollution.

42

On Tariff Design * Since the experience of Corrientes shows that

* Consider the preparation of a standard tariff private financing is hard to come by forsystem (as was done in the case of electricity private concessionaires and that, for now,system (s was the ES to rvse an possibriciy many provincial governments will have littledistribution by the ES) to revise and possibly choice but to finance a share of the investmentsimplify the intricate fixed tariff for unmetered roects of the rivate operators who cannotconsumption and to increase the incentive for brow fo the ban pectors to minnotlarge residential users to accept metering. borrow from the banking sector; to minimize

the short term financing requirement to the* Revise the fixed part component of the two- public sector, the governments may want to

part tariff to make sure that it reflects the consider alternative arrangements such asactual connection cost and not just some BOTs for specific expansion projects inapproximation of the ability to pay that is combination with partial concessions orunrelated to the cost of delivering the service. management contracts.

* Ensure that the revisions address the social * Other ways of reducing the limitations onconcerns (such as the need of the poor and financing include attempts to try to concessionenvironmental problems) as well as the only specific areas or to bundle W&S assetsincentives of the concessionaire and ensure with other assets such as electricity to reducethat new and old users are all paying the long- the risks faced or perceived by privateterm marginal costs needed to cover the investors.expansion needs of the sector.

* Whatever their form, ensure that contracts are

On Contract Selection goal-oriented rather than process-oriented andthat they do not have mutually exclusive

* Adjust the contract type to fit the preferences requirements.

of the government or to fit what the private * Specify clear rules for contract renegotiation.sector is willing to accept; managementcontract with a minimum level of capital On Institutional Reformcommitment may be the best option that canconsidered in the short run in some provinces. * Streamline the responsibilities for the sector at

the national level.

43

Chapter 4: DEVELOPING A PROVINCIAL REGULATORY CAPACITY

4.1 The previous two chapters showed that instance providing more services to more peoplemany provincial governments are trying to retreat with about 50% less staff than Obras Sanitarias.from the role of operator of electricity and W&Sservices wherever possible. As they do this, they 4.4 The pressure on the reformingare progressively placing more emphasis on their governments is however often strong to avoidrole as regulator of the private monopolies. This dismissal and to maintain a tradition of usingshift is not easy and requires a strategy for infrastructure related services as a source ofdefining and implementing regulatory frameworks. employment. The main temptation has been toAs in the case of other countries with federal place the employees dismissed by the privatesystems, a special challenge in Argentina is the operator in the regulatory agency. In someestablishment of institutional arrangements that provinces the preliminary number of staff in thereflect the proper roles of national and provincial regulating agencies was expected to be larger thantiers of government. the number of regulated. The risk with this

strategy is that the regulator becomes a source of4.2 The importance of these institutions is inefficiency and ends up hurting the interest ofbeing underestimated in some provinces. Their users rather than protecting them.role is also sometimes misunderstood as revealedby some of the decisions already made by some 4.5 Potential investors in infrastructureprovincial governments. Luckily few decisions activities are aware of this risk and many othershave been sealed in stone and it may be useful for including interference with pricing of thesome provincial authorities to have some insights vulnerability of their usually large, long-term andon some of the costs and benefits of their highly-specific investments once they have beendecisions and to provide with adjustment options made. To minimize these risks, governments needin cases in which they end up concluding their to break with their past and be able to commitinitial strategy may have been to hasty. The credibly to regulatory policies to provide investorschapter provides a framework for thinking about with assurance of a fair return on their investment.the institutional requirements of the This condition is necessary both for attractingimplementation of the regulatory arrangements. initial investment at reasonable cost and forThe chapter focuses on the issues that were encouraging efficient operation of that investmentconsidered the most important in the discussion once it has been made.held during the visits to the provinces.

4.6 The development of a regulatoryBreaking with the Past framework that limits discretion in regulatory

rules discussed in the previous two chapters is4.3 The water and electricity sector in only part of the strategy needed to show aArgentina have traditionally been significant govemment's commitment to reform. The neededsources of employment in many provinces. It is complementary element of this strategy is to placeoften a major factor in the high costs of the responsibility for the exercise of residualservices. It is also one of the first changes a discretion in an agency with the requisite expertiseprivate operator wants to bring to the way the and independence. "Independence" in this contextservice is being delivered. Aguas Argentinas is for includes measures to protect the regulatory

44

decision-making process from being "captured" by may be required. Moreover, if the regulatoryshort-term political pressures reflecting the needs entity is intended to be independent, decision-of special interest groups (often pushing makers must be able to resist improperregulators to keep tariff lower than desirable) and inducements or pressures from regulated firms,by the concessionaire (often pushing regulators to political authorities, and other interest groups.keep tariff higher than desirable). For manyprovinces this may the most difficult break fromthe past. It includes the adoption of a transparent Box 4.1 How to Appoint Regulators?mechanism for the selection of autonomous Assuming that the tradition of appointingregulators (see Box 4.1). Regulatory Commissions (with 3-5 members rather than a

single regulator) remains the preferred options in

Constraints and Trade-Offs Argentina (and there is evidence that this is the right wayto go), the selection of the members is the first severepolitical challenge provincial authorities will have to face.

4.7 The most appropriate means of Here the main lesson of experience is that elections are notimplementing this strategy will vary from province appropriate: they would reinforce short-term politicalto province. In provinces with stable political and perspectives when longer term perspectives required.legal institutions and an established reputation for Best practice includes the following steps:treating private firms fairly, perceptions of lower * prescribe professional qualifications for appointees;regulatory risk may reduce investor's demands forsafeguards against the misuse of regulatory . involve both the Executive and Legislative Branches indiscretion, such as rigid and specific rules and the appointment process;independent agencies. More flexible rules will, in . provide regulators with security of tenure, subject toturn, enhance regulators' capacity to pursue removal only in case of proven misbehaviorefficiency goals and to adapt regulation to * exempt agencies from civil service rules that limit salarieschanging economic and technological conditions. below those required to reeniut qualified staff,

. provide agencies with funding from industry levies and/or4.8 In provinces with less conducive user fees to reduce reliance on politically-directedconditions, the trade-offs can be stark; investors budgetary allocations.respond to increased risks by insisting on greater Measures to provide regulators with a degree ofsafeguards or demanding higher returns on capital. insulation from the political process have to be reconciledThe less comfort provided by the regulatory with the need to ensure regulators remain accountable forsystem, the higher the risk premium required to their actions. While the means of striking the appropriateattract investment. Another trade- off also arises: balance between autonomy and accountability will varythe lower the confidence in the regulator's from context to context, common measures include:independence, the greater the emphasis on . provisions permitting removal of regulators in the casespecific, rigid rules and hence the constrains on of proven misbehavior;using regulation to pursue efficiency goals. . prescribing particular duties and obligations as clearly

as possible in law, and providing an effective appeal

4.9 Dealing with limited regulatory process, whether involving the courts or anothercapacity. The main problem many of the forum; andprovinces will face is their constrained regulatory * mandate high standards of transparency in regulatorycapacity. Regulation of infrastructure industries is proceedings.a complex and demanding task. Depending on thedetail of the regulatory regime, staff withspecialized economic, financial, and legal skills

45

4.10 Such resources are scarce in Argentina's Countries which use sector regulators for energyprovinces and often have a high opportunity cost. include Colombia, Hungary, and the national-The principal concern is the relative scarcity (and level regulators in the United States and Canada.high opportunity costs) of qualified regulatory Sectoral regulators are being established inpersonnel at lower tiers of government, with this Russia and are being advocated as a reformconcern exacerbated by the greater risk of option in the United Kingdom.regulatory capture flowing from closer proximity Multisectoral. A single regulatory agency isto consumers (voters) and firms. These concerns responsible for all or most infrastructure sectors.affect the design of regulatory arrangements and Examples include state-level regulators in theof strategies to strengthen the regulatory capacity U.S., Canada and Australia, as well as national-of decentralized regulators. I level arrangements being implemented in Jamaica

4.11 In anenionen fimtdeand Bolivia. In the provincial Argentinean4. . n anenvironmentoflimitedregulatorycontext this would mean a single regulatory

capacity, potential investors tend to be especially agency would cover W&S and electnicitysensitive to discretion in regulatory arrangements,and have a strong preference for detailed and rigid distrbutionrules and standards. To the extent that discretionis tolerated -- and as a practical matter, it is 4. 13 The multisectoral approach is probably theis t a - d a cmost desirable in the context of Argentina'svirtually impossible to remove all discretion -- apremium will be placed on measures that E iprovices needs and lmited regulatory capacities.safeguard the independence of the regulator. ustablshng regulatory wth a broad base offers aThese measures may range from procedures agenumber of advantages over industry-specificaffecting the appointment process and security of agencies. These advantages are even more

tenure~ ~ to poeue.nurn rnprn important when provincial regulators are involved.tenure to procedures ensuring transparent Thadntgsicuedecision-making and the review of decisions. Tothe extent that these conditions are not met, . Opportunities to share regulatory resources --investors will demand higher rates of return to for example, economists, financial analysts, andoffset the regulatory risks they perceive. lawyers can work across sectors, and

admninistrative personnel can be shared. ThisSectoral Breadth of Authority reduces operational costs and means lower tariffs

and/or lower burden on taxpayers.4. 12 Entities charged with regulating * Greater resistance to industry capture -- theinfrastructure can be organized on one of three agency's broader constituency weakens themain bases, depending on the sectoral breadth of influence of any one industry.their authority.

I* Greater resistance to improper political. Indutryspecc S t a s ainterference -- the broader constituency and

responsible for gas, water, electnrcity, rail, and so higher poltical profile increase the stakes ofon. This has been Argentina's approach at the inappropriate political intervention.National level.

. Application of consistent approaches across.Sector-specific. Separate agencies are industries -- this has two main benefits. First, it

responsible for sectors comprising a group of contributes to greater predictablity (and hencerelated industries, such as an energy regulator for reduced risk) for investors. Second, it reduceselectricity and gas, a transport regulator for rail, the risk of market distortions that can arise when

roads, and ports and a communications regulator services that are subject to substitute competitionfor teleconmunications, post, and broadcasting.

46

-- such as gas and electricity or different departments. This approach has been adopted bytransport modes --are subject to inconsistent many state-level regulators in the U.S. and isregulatory treatment. being emulated in Bolivia.

Easy translation of lessons and experience gained Iin one sector to other sectors -- this can be suRepreetacparticularly important in relation to the design ofefficient tariff structures and some of the new 4.16 In view of the high degree ofregulatory challenges associated with managimg decentralization of the W&S and electricity

competition in network services such as distribution services, the degree of localtelecommunications, electricity and gas. representation is an important question to sort

4. 14 Implementation strategies. Three broad out. It is important that consumers know thatstrategies can be adoptedfr estabishtheir interests are being protected, and can refer

Mategiesectran regulatorypfrmeworkes.aFirstn a problems to the regulator when required. But thismultisectoral framework can be established at the does not mean that the regulator needs a

outset,witheachsectoralframn br h withbinhed re gie representative office in each municipality -- otheroutset, with each sector brought within the regime cutiswt ogeprec nti rahvcountries with Iong experience In this area haveat the time of or before privatization. This will c g usually be the preferred approach, and has been chosenadopted in countries including Bolivia. Second, ifan industry-specific regulator already exists, new * use of toll free telephone numbers forindustries may be brought within its jurisdiction reporting concerns,rather than creating additional entities. The ..

... . ~~~~~~~* use of a program of regular visits tofeasibility of this approach will depend on how municipalities to receive concerns" explaineasii' the structure and operation of the initial p aitstitution can be modified to meet a broader poice and whe taifs et. ar en

iiistitulion can be modified to meet a broaderlreviewed, seek submissions through publicmandate. Third, regulators can be established onan industry-specific basis but consolidated overtime through mergers. This happened in many . if justified by the geography (and politics) of astates in the U.S. and pressure is mounting to particular province, establish one or twoierge industry-specific regulators in the U.K., regional offices serving a large number of

Ar-gentina and Chile. This strategy has a number municipalities.of weaknesses, however, including the likelihoodthat existing entities will resist merger and that the The possible establishment of regional orbenefits of a broader perspective will not be municipal offices also underlines the benefits ofavailable during the critical early phases of a new multisectoral agency or at least sharing offices andregLilatory system. staffs between regulators.

.4. 15 One of the main challenges in establishing Minimization of Regulatory Demands"iultisectoral agencies is to ensure that specializedknowledge or understanding of particular 4.17 If regulatory capacity is limited, thisMndustries is not unduly sacrificed. This challenge limitation must be taken into account in designingCean. be addressed by ensuring that internal all aspects of the regulatory system. RegulatoryWganizational arrangements foster the discretion can be limited and residual discretion

development of industry-specific expertise, such can be tempered by clearly articulated criteria.as bv establishing strong sector-specific Administrative procedures intended to make

47

regulatory decision-making transparent should 4.21 For example, price regulation in the U.S.avoid excessive legalism, a criticism often leveled involves rate of return methodology and almostat regulatory processes in the U.S. but also a continuous public hearings undertaken in a verytendency of many involved in the design of legalistic environment -- hence, large number ofregulation in Argentina. The design of staff consumed with procedural matters andenvironmental regulation for instance is a major detailed evaluation of accounting data from firmsheadache for any water company operator. In contrast, price cap methodology in the UK

requires rates to be reviewed only once every4.18 The burden of supervising compliance with three to five years, and processes for publicrules can be reduced by ensuring that penalties are consultation are less formal and legalistic.clearly defined and set at a level that will provideappropriate incentives for firms. It may also be 4.22 Quality regulation provides anotherfeasible to enhance the role of consumers in example. The "old model" involved regulatorsmonitoring firm behavior by providing incentives acting as policemen, actively supervising serviceto report non-compliance. For example, in delivery and searching out infractions. Inaddition to fines, consumers may receive rebates if contrast, the "new model" places more emphasisa firm has failed to meet clearly defined on deterrence (large penalties) and incentives forperformance obligations. consumers to inform the regulator of non-

compliance (e.g., rebates from bill when service4.19 Contracting-out tasks. Another way to not available or falls below agreed parameters)reduce demands on regulators is to allow them to (see Box 4.2). These issues are determined bycontract-out particular tasks. For example, some regulatory framework, which agency usually mustof the analytical work associated with price take as given -- hence, importance of thinkingregulation and tariff design in the water sector through these issues before privatization!could be contracted-out to consultants. Externalarbitrators could be used to settle certain types of 4.23 Large staff is not necessary. For instance,regulatory disputes. Similarly, some regulatory In Colombia (which is larger than many oftasks may be delegated to national agencies where Argentina's provinces), the new regulator forthe capacity is demonstrated and the incentive to electricity and gas throughout the country has justensure an effective coordination with the functions 18 staff, and does not anticipate expansion beyondunder their responsibilities is obvious. In each 36. Sector-specific national regulators incase it will be important to ensure that the countries the size of Argentina often have 40 orproposed arrangements do not raise concerns over less staff, and multisectoral agencies in the U.S. --capture or otherwise weaken the legitimacy or despite the demands of burdensome rate-makingeffectiveness of regulatory decision-making. and legal processes -- often have around 100 staff

to cover a broad range of sectors.Size of Agency Staff

4.24 It is also important to emphasize the costs4.20 There is no "magic number", and no and risks involved in erring on the side of havingobjective indicator based on size of firm, turnover "too many" regulatory staff:of industry or number of consumers. It all of extra agency staffdepends on the specific tasks assigned to theregulator under the particular regulatory regime. * larger number of regulators will likely increase

demands on firm, so that its staff must expandaccording, with these costs ultimately reflectedin higher prices;

48

* risk that larger staff will become toointerventionist to justify their existence -- Box 4.2 Consumer Representation"micro-management" of industry eliminates There are two main models for ensuring views ofbenefits of private participation; concerns in consumers are taken into account: (i) have a representativethis area may deter potential investors or lead of consumers on commission, together with representativesthem to reduce the value of their offer of producers and other interests; and (ii) staff commission!accordingly; and with professionals (economists, lawyers, etc.), but ensure!decision-making process gives opportunities for interest

* large staffs make it more difficult to pay higher groups and interested individuals to present their views.salaries to attract and retain well-qualified Both models have precedents around the world.staff But model (ii) has several advantages over model (i), and

is the model of choice in the UK and most U.S. agencies.4.25 The main message here is that quality is Advantages include:niore important than quantity. It is usually easier . decision-making and balancing of interests is doneto hire staff than to fire them. So the preferred more transparently, rather than through internalstrategy should be to begin with a small number of negotiation;-well-qualified professionals, engage short-term * encourages decision-makers to take a broad. long-termnconsultants for specific problems, and only expand perspective, rather than just acting as an advocate for fstaff when clear that they are required to perform one group;specific and ongoing duties. * ensures commission is qualified to deal with often

technically-complex issues involved; andStrengthening Regulatory Capacity avoids disputes over composition of commission -- in

reality, there are diffuse consumer interests (e.g., lowincome residential; middle income residential; high

4.26 Even where regulatory arrangements have volume industrial consumers; small-to-mediumbeen designed to reduce regulatory demands, commercial users). No one individual can expect tonascent regulators will typically require assistance faithfully represent all these views, and appointment ofin a number of areas, including training in specific a large number of consumer interests requires theregulatory functions. This type of assistance will appointment of an equal number of producer interests

for balance -- the result becomes a very largeusually exhibit significant scale economies, as commission, with consequent delays, poorItraining programs and other materials may be accountability for decisions, difficulty for investors toshared by several decentralized regulators. A predict policy directions, etc.number of strategies exist for delivering assistanceof this kind.

4.28 Mutual Assistance Through a4.27 Assistance Through National-level Regulatory Association. An alternative orAgencies. A national policy favoring complementary strategy is to establish andecentralization of regulatory responsibility may association of regulators. Sub-national (andbe complemented by technical assistance national) regulators can interact in this forum andchanneled through a national-level agency. For share lessons of experience. The association mayexample, such an agency could organize training also reinforce professional norms (and henceprograms for regulators, provide model contracts reduce vulnerability to improper influences) whilefor concessions, or provide technical advice on providing a group of peers to respond to charges ofspecific regulatory problems. It may also act as a improper political interference. It could also providefocal point for disseminating lessons of experience a framework for the joint development of trainingamong sub-national regulators. programs and research on common regulatory

problems. The National Association of Regulatory

49

Utility Commissioners in the U.S. provides a possible concession in Buenos Aires. 34 This approach hasmodel and this is what Argentina's electricity the benefit of improving the transparency of costsregulators are trying to achieve as explained in associated with regulation, but may also increasechapter 2 (para. 2.28). the costs and complexity of administering the

scheme.4.29 Twinning Arrangements. A third vehiclefor supporting nascent regulators is establishing 4.32 The second approach -- imposing the levy"twinning" relationships between sub-national on regulated firms -- is the most common, and isregulators and more experienced regulators, the model used for most regulators in the U.S. andwhether from foreign or national jurisdictions. Argentina and for regulators in the UK,Arrangements of this kind can provide a source of Venezuela, Peru and Colombia. The manner inongoing institutional support, technical advice and which the levy is passed on to consumers willexchange and training opportunities. A growing depend in part on the form of price regulation.number of U.S. and other OECD regulators have Under traditional rate of return regulation, thisexperience in participating in these relationships cost will often be treated as an element ofwith regulators in reforming and developing operating costs and hence passed on to consumerseconomies. automatically through the tariff schedule approved

by the regulator. Under price cap regulation,Funding the Regulatory Agency accounting for individual cost items is less

important, and tariff increases do not4.30 A critical question in establishing automatically flow from changes in a firm's costautonomous agencies is how to strike the balance structure. As price cap schemes are neverthelessbetween autonomy and accountability, including in intended to allow efficient firms to make arelation to funding arrangements. Relying on reasonable rate of return, the overall result maybudgetary transfers is often viewed as a threat to not be that different from under rate of returnthe independence of the regulators. Levies on regulation. However, the risks implicit in priceregulated firms tend to be the most important cap regulation fall on the regulated firm, ratheralternative source to funding to general budget than consumers, which may make a firm operatingreceipts and could be considered by reforming under price cap regulation more sensitive to theprovincial governments. Actual levies are level of the industry levy.determined annually based on budget proposalssubmitted by the agencies Some of the main 4.33 What kind of fees? The design of the feeimplementation issues are discussed below. requires a few decisions. Fees can be levied as a

percentage of the industry's gross revenue and4.31 Imposing the Levy on Firms or on paid every six months or so. They can also beConsumers? In principle, regulatory costs could levied as a flat fee to be paid early on in the year.be imposed on consumers either directly, such as To avoid the risks of excessive burden, athrough a separately identified fee as part of their maximum level of industry levies can be set in law.utility bill, or indirectly, such as where the levy is In Peru, for example, sectoral laws provide thatpaid by regulated firms but those firms pass the the maximum levy for water firms shall be 2% ofcosts on to consumers in the form of higher industry gross revenues. In Colombia, theprices. The first approach remains relatively

unusual, with examples including the regulator 34 In Buenos Aires, the water regulator applies a levy ofresponsible for overseeing the private water 2.67% on consumers' water bills. In the District of

Columbia, a fee of $0.03 per month for residentialcable subscribers was introduced in November 1994.

50

electricity law provides that levies on regulated * sources of mutual support if concerns overfirms cannot exceed 1% of specified components improper political intervention (i.e., akin to aof the entities' costs. But while a maximum levy "trade union" for regulators); andmay provide regulated firms with some degree of .jointly supporting an effective system ofprotection from imposts they consider excessive, yardstick competition between provinces.and set a cap on de facto taxation, they do notprovide agencies with any protection from 4.36 In the U.S., State and Federal regulatorsarbitrary political interference. To achieve this have combined to form the National Associationgoal, a minimum levy would need to be specified for Regulatory Utility Commissioners (NARUC)

in law or some other instrument that cannot which performs these roles, has specialreadily be amended or revoked by government, committees for each sector, and jointly finances a

4.34 What kind of fee structure? Where the National Regulatory Research Institute (NRRI) at4.34 What kinod ofifestruon cture? Where the Ohio State University to undertake research andlevy is imposed directly on consumers, or is training on behalf of its members. A similar modelimposed on firms but tariff regulation expicitly may have many benefits in Argentina. Argentina;directs the incidence of the regulatory fee electricity regulators are hoping to achievecomponent of the firm's costs, questions arise as something similar with the agreement signed into the design of the most appropriate structure of Aprilg gimilar got be surprisig inlevies. For example, if the regulatory fee is April 1996 and it would not be surprising if thesignificant, redistributional objectives might be water sector will follow soon with a similarpursued by exempting some categories of users gfrom the regulatory fee. More generally,efficiency objectives will generally be advanced by summing upimposing the fee on tariff elements with least

elastic demand, such as a connection fee, rather 4.37 Selectinguthe most appropriate design for athan a demand fee. The costs and benefits of Provincial regulatory entity will call for complexadopting more elaborate schemes of this kind need policy judgments, often made in a sensitiveto be weighed carefuilly in the context of particular political environment. This choice cannot be

implementation environments. made in the abstract or by appeal to generalslogans. Constrained regulatory capacity is of

Organizing Interprovincial Cooperation on special concern in a reforming country such asRegulatoryMatters Argentina. This constraint must be given due

weight in framing any privatization and

4.35 Opportunities include: decentralization strategy, and when consideringthe detail of regulatory and institutional

* realizing economies of scale in training arrangements. In particular, provincial

programs; governments may need to find ways to resist the

* realizing economies of scale in comrnissioned pressures that can arise for each industry to have aresearch on specific regulatory issues; separate industry-specific regulatory agency.

Thought will also need to be given to strategies* informal exchanges on lessons of experience in for strengthening regulatory capacity at all tiers of

dealing with particular issues; government.

Repo rt No

Ty Pe~~~r

AR~~~~~~~~~~~~