Regulation policies concerning natural monopolies in developing

ST/ESA/1999/DP.8DESA Discussion Paper No. 8

Regulation policiesconcerning naturalmonopolies indeveloping andtransition economies

S. Ran Kim and A. HornMarch 1999

United Nations

DESA Discussion Paper Series

DESA Discussion Papers are preliminary documents circulated in alimited number of copies and posted on the DESA web sitehttp://www.un.org/esa/papers.htm to stimulate discussion andcritical comment. This paper has not been formally edited and thedesignations and terminology used do not imply the expression of anyopinion whatsoever on the part of the United Nations Secretariat.Citations should refer to a “Discussion Paper of the United NationsDepartment of Economic and Social Affairs.”

S. Ran Kim and A. Horn

Ms. S. Ran Kim is associate expert and Mr. A. Horn is Deputy Director ofthe Division for Public Economics and Public Administration, UnitedNations Department of Economic and Social Affairs, New York. We arevery much indebted to valuable comments and suggestions from Mr. TonyBennett. Comments should be addressed to the authors, c/o Division forPublic Economics and Public Administration, Rm. DC1-900, UnitedNations, New York, N.Y. 10017, or by e-mail to [email protected] copies of the paper are available from the same address.

Authorized for distribution by:

Guido BertucciDirectorDivision for Public Economics and Public AdministrationRoom DC1-928United NationsNew York, NY 10017Phone: (212) 963-5859/Fax: (212) 963-9681Email: [email protected]

United Nations

Department of Economic and Social Affairs

Abstract

Network industries are often organized as vertically integrated public monopolies. Recent trends indicatethe participation of the private sector. Developing and transition economies need to establish adequateregulatory policies and institutions to provide incentives for private sector participation and to protectpublic interests. New regulatory policies entail the creation of market competition in such industries oralternatively the creation of competition for the market. Natural monopoly sector privatization is arelatively new and still-evolving field, and it would be premature to venture definitive conclusions as to the“best practice” privatization and regulation models for natural monopolies. Nevertheless, we will offersome recommendations concerning natural monopoly privatization and regulation.

Introduction

Recently, there has been a significanttransformation in the style of natural monopolyregulation policies away from the previousalmost exclusive reliance on public ownership.Once public ownership was hailed as a “reform”,but now privatization has become a “reform”.“Reform” today means deregulation, competitionand privatization.

Privatization and restructuring ofnetwork industries traditionally viewed asnatural monopolies have been gaining groundrapidly around the world since the early 1980s --implying a radical shift in the focus of stateintervention and a re-evaluation of the State’srole even as a provider of core public services.

Why then this change? And to whatextent did this change take place, in particular intransition and developing economies? What are

the empirical trends of natural monopoly sectordevelopment and regulation in these countries?

Where are the current natural monopolyregulation models in these countries heading?What are their recent experiences in the domainof natural monopoly regulation? After adiscussion of what natural monopoly means andthe major issues that are currently being debatedin the area of natural monopoly regulation, wewill examine the changes taking place in variousnetwork industries.

Natural monopoly sector privatization isa relatively new and still-evolving field, and itwould be premature to venture definitiveconclusions as to the “best practice” privatizationand regulation models for natural monopolies.Nevertheless, we will offer somerecommendations concerning natural monopolyprivatization and regulation.

Defining Natural Monopoly and Its Current RegulationPolicy Agenda

Many network industries have beenpredominantly provided by a verticallyintegrated, often public, monopoly. However,since the early 1980s the paradigm of publicmonopoly has been losing ground with thesteady breakup of the activities traditionallyregarded as natural monopolies(demonopolization) due to globalization ofmarkets and technological progress. At the sametime, growing dissatisfaction with publicenterprise performance, ever-tighteninggovernment budgets, and the explosion ofinvestment needs in utility and other networkindustries worldwide have caused policymakersto turn increasingly to private sectorparticipation.

This often intertwined reform process ofprivatization and demonopolization of theseindustries initiated in the developed countries isalso sweeping across the developing countries.In the wake of this change, the current agenda ofregulatory policy concerning natural monopolyis not limited to the traditional price and entryregulation issues. Rather, it includes the issuesrelated to the design of regulatory institutionsaccompanying the restructuring, privatization,and expansion of competition in the area

formerly occupied by regulated, often public,monopolies.

The concept of natural monopoly,traditional regulatory practice andrationale

A natural monopoly exists when economies ofscale are so substantial that a single firm canproduce total business output at a lower unitcost, and thus more efficiently than two or morefirms (Sherer 1980). In effect, the long-runaverage costs are falling over such a wide rangeof production rates (relative to demand) that onlyone firm can survive in such an industry. A morespecific criterion is the subadditivity of the costfunction.

Natural monopoly gives rise to apotential conflict between cost efficiency andcompetition, with an increased number ofcompetitors leading to some loss of scaleefficiencies. The typically quoted examples ofnatural monopoly are utilities (electricity,telecommunication, water, gas, and oil),transport (railways), with natural monopolyelements being centred on networks (Yarrow1994).

2 DESA Discussion Paper No. 8

An electric company is a classicexample of a natural monopoly, wherecompetition may lead to an inefficient marketoutcome. Once the huge fixed cost involved withpower generation and power lines are paid, eachadditional unit of electricity costs very little.Having two electric companies split electricityproduction, each with its own power source andpower lines, would lead to a near doubling ofprice, because of low marginal costs, high sunkcosts and declining average costs.

Natural monopoly thus poses thedifficult dilemma of how to organize theseindustries so as to gain the advantages ofproduction by a single firm, while minimizing allthe vices resulting from non-competitivemarkets.1

Traditionally, countries around theworld, assuming the “inevitability” ofmonopolization, either regulated privateenterprises or nationalized natural monopolies inorder to deal with this dilemma.

A natural monopoly situation usuallyarises when there are large fixed costs and smallmarginal costs. The existence of a naturalmonopoly gives rise to the following problem:Allowing a natural monopolist to set themonopoly price is undesirable due to the Paretoinefficiency, and forcing the natural monopoly tosell at the efficient price (i.e., marginal cost-based price) is infeasible due to negative profits.The solution to this problem was then to let thegovernment operate the service, for example, atprice equal to marginal cost and to provide alump-sum subsidy to keep the firm in operation.This practice rests on the assumption that theimposition of public interest prices and standardsmay be achieved more effectively by the flexibledecision-making inherent in the publicownership framework—considerable internaldiscretion, subject only to politicalaccountability—than by legal controls of privatefirms (Ogus 1994, pp. 267-68). Otherwise,regulation of private monopolists has usuallyinvolved some form of price regulation and/orentry and quality regulation. 1 These vices range from ‘deadweight welfare loss’due to allocative inefficiency, productive inefficiency (or x-inefficiency) due to lack of competitive pressures, increasedpossibility of collusion among firms, increased possibility of‘predatory pricing’ or ‘pre-emptive investments’ and other‘wasteful’ behaviour to increased possibility of exploitationof consumers and of input suppliers by the dominant firms(Chang 1997, pp. 707-708).

These regulatory practices weretheoretically underpinned by the market failureargument, which provided the central economicargument for state intervention in industries withnatural monopoly characteristics.

Alongside other conditions such aspublic goods, positive and negative externalities,incomplete markets and imperfect or asymmetricinformation, natural monopoly is an importantmarket failure situation (under which a marketeconomy fails to allocate resources efficiently)that warrants regulation and nationalization.

In fact, it is argued that the most seriousmarket failure problems are likely to occur innetwork industries with natural monopolycharacteristics. According to Yarrow (1994), thisis because natural monopoly is combined withhigh-entry barriers. These industries are typicallycapital-intensive and require significantinvestments in long-lived, sunk capital facilities.Most assets are specific and durable, giving riseto high-entry barriers via extensive sunk costs.At the same time, the economies of scale in someindustries such as water distribution or electricityare so great that the largest firm with the lowestcosts could drive all other competitors out of themarket.

It is important to note that regulation ofnatural monopolies also occurs for reasons otherthan market failure (generally considered a staticefficiency problem). In fact, many real liferegulations have been motivated by the concernfor dynamic efficiency, distributionalconsiderations and other considerations,including even “moral” considerations—such asfairness.

In particular for developing countries,dynamic efficiency (or in other words,developmental) objectives such as growth areoften more important than static efficiency. Themost important dynamic efficiency considerationis, as Bradburd (1992) points out, whether anunregulated private monopoly will make theinvestments necessary to offer the quality ofservice appropriate to the country’s changingneeds over time. Natural monopolies’ servicesare an important part of a nation’s infrastructure,and if they are suboptimally provided, this canbe an impediment to growth.

Thus, the new regulatory reform shouldgive adequate attention to considerations of“dynamic” efficiency. Some countries conduct

3 Regulation Policies Concerning Natural Monopolies

their deregulation-based reform purely in termsof static efficiency, and the impacts of regulatoryreform on productivity and growth are not dulyconsidered when reforming the existingmonopoly regulation policy. This is a highlyinadequate approach, as Chang (1997) pointsout, as higher static efficiency will notnecessarily lead to higher dynamic efficiency. Inaddition, removing “distortions” in more, but notall, markets does not necessarily improve eventhe static efficiency of the economy. Schumpeter(1987) argued that monopoly rents provide theincentive to innovate and, in the modern age oflarge-scale R&D, the resources to innovate. Ifthis is true, there may even be trade-off betweenstatic and dynamic efficiencies. If the regulatoryreform involves reductions in market power andthe associated monopoly rents (e.g., byintensifying anti-trust regulation), the rate ofinnovation and productivity growth may beadversely affected.

Regulation practice is driven not onlyby normative considerations of reducing andcontrolling rent-seeking behaviour. The positivetheory of regulation, based on public choicetheory, treats the existence and forms ofregulation as responses to the demands ofpoliticians and other interest groups.

In summary, the traditional rationalesand a wide range of (non-static efficiency) issuesthat traditionally belong to the realm of naturalmonopoly regulation policy may still remainvalid and require adequate attention when“reforming” the existing regulatory regimes.While reforms may be necessary to makeservices more efficient and economical, the usualpublic service raison d’être of many naturalmonopoly industries also remains essential.Particularly in the developing world context, it isimportant to keep in mind that the ultimateobjective of these industries is sufficient andsustainable provision of their services.

Forces of change, new regulatoryagenda and theoreticalalternatives

Recently, new developments such astechnological progress, which offer means ofcontesting a monopoly, have fundamentallychallenged the traditional regulatory practices

based on the concept of natural monopoly. Thesteady breakup of “intrinsically monopolistic”network industries into separate elements haslargely obviated the justification for theexistence of large, vertically integratedmonopolies.

There are increasing doubts whethersome of the industries traditionally regulated doin fact have the structural characteristics of anatural monopoly. Many traditional naturalmonopolies have been shown to be less naturallymonopolistic than was once thought to be thecase.

The degree of natural monopoly ofmany industries has also been drasticallyreduced, due to technological progress andglobalization of markets, though not eliminatedentirely (World Bank 1997). Some even arguethat there is nothing “natural” about “naturalmonopolies”, challenging the very concept ofnatural monopoly (see for example Becker1997).

The most important challenge istechnological progress, which changes the costcurves, hence enabling countries to re-examinethe hitherto characteristic forms of naturalmonopoly regulation, i.e., price and entryregulation, based on the concept of naturalmonopoly.

New technologies evolved that areefficient at much lower levels of output thanolder methods of production. These havesubstantially reduced economies of scale andbarriers to entry in many sectors, making at leastsome degree of competition for many naturalmonopolies a real possibility. Development ofnew technologies such as wireless telephony andoptic-fiber cable has created new scope forcompetition even with regard to basic linenetworks. In electricity, with combined cycleturbine generators, we have a low-capital-costsource of power, which cancels out economies ofscale in generation and voids any argument thatelectricity generation is a natural monopoly. As aresult, even in some traditional naturalmonopolies such as telecommunications (e.g.,long-distance and wireless telephony networks)and electricity generation, market competitionhas become both possible and desirable.

The possibility of extending the marketsize due to globalization has undermined theeconomic rationale of monopoly retention


policy. As Yarrow (1994) points out, whether ornot an industry is a natural monopoly dependsupon technology/costs and demand. Thus,natural monopolies can disappear or emerge asdemand expands or contracts, even if productionconditions do not change. According to Becker(1997), the growth of global competition impliesthat when large-scale production is mostefficient, companies in small nations are nolonger restricted to the inefficiently small scaleof their limited domestic market. They canincrease production enormously by operating inseveral nations.

What then are the consequences ofthese new developments for the naturalmonopoly regulatory policy debate? Which newissues and changed regulatory demands are thenbrought into the domain of regulation policyconcerning natural monopolies?

The focus of regulatory policyconcerning natural monopoly has clearly shiftedwith the evolution of technology andglobalization of markets, which led to a steadybreakup of natural monopoly and made morecompetition technically feasible. Instead ofmerely focusing on problems surrounding“inevitable” monopolization such as the pricingproblem, the current regulation policy henceencompasses, above all, issues related to thedesign of regulatory policy accompanying therestructuring, privatization, and expansion ofcompetition into the area formerly occupied bylegal monopolies. In particular, the issue of howto replace regulation with competition, which isdeemed as the best regulator, now occupies acentral place on the current agenda of naturalmonopoly regulation.

Part of the debate over regulationconcerns the limits of natural monopoly in theface of technological change. The verticallyintegrated, often public, monopolies have nowbeen shown to be no longer monolithic entities.Rather, they encompass services that arearguably natural monopolies as well as servicesthat are potentially competitive but need accessto bottleneck monopoly or certain essentialfacilities to make competition in these supplysegments feasible (Joskow 1998).2 In particular, 2 Generally speaking, physical infrastructures tendto have monopolistic characteristics, and services competitiveones. See Guislain (1997, pp. 212-14) and Plane (1998, p. 14)on how to “unbundle” different sectors into their componentactivities. The organizational and institutional reform is then

“unbundling” of monopolistic firms isconsidered as one of the most exciting ways toaccelerate competition. Unbundling isolatesresidual sunk-cost facilities (e.g., the local loopin local telecom), leaving the contestable part ofthe industry under the control of market forces(see Teece 1995).

In practice, determining where theboundaries of natural monopoly is a difficultexercise, which requires detailed information onwhat may be quite complex cost conditions(Yarrow 1994). Nevertheless, a consensus hasemerged for the need to revive the rules ofmarket competition, whenever high fixed-costactivities cease to justify the presence of a singlemonopoly firm.

Consequently, out of this changedcontext, some issues have emerged as the new,important regulatory policy issues on the agenda.These are for example “unbundling” of a singlemonopoly, restructuring, and scaling back ofmonopoly protection through demonopolizationto remove artificial monopoly privileges, whilelimiting legal, generally public monopolyprotection to those aspects of the activity thatjustify the “natural” monopoly.

Additional recent developments on thetheoretical front have enforced this embrace ofthe competitive model, as the right way toorganize many network industries previouslyviewed as natural monopoly industries. Scholarlywork has begun to emphasize that naturalmonopolies do not necessarily have to beregulated, since there are alternative ways togenerate competition and discipline the firms,even if a natural monopoly structure existswithin a market (Brauetigam 1989).

According to this argument, there existthe following theoretical alternatives:

Competition for the market: Onepossibility is to retain the monopoly but to createcompetition between firms for the right ofexclusive supply over a limited period, namely afranchise solution. This has been formalized asDemsetz-competition. The essential idea is thatsuch competition for the market (i.e., the right tobe the natural monopolist) may be an adequatesubstitute under some circumstances, wherecompetition is not possible within the market.

based on an economic analysis aimed at identifying the linksin the technical chain where the cost function is sub-additiveand the market is not contestable.


The outcome of Demsetz competition is in effecta contract between a franchiser (e.g., agovernmental authority) and a franchisee.Monopoly franchises could be auctioned off tothe bidder offering the most attractive terms, forexample, the lowest price to consumers.Franchising schemes also may avoid pitfallsassociated with traditional regulation of suchindustries or with their nationalization. Wherecompetition cannot be introduced in the market,as tends to be the case for water supply, forexample, it should at least be introduced for themarket. Properly structured tenders or auctionswill allow the government to extract part of themonopoly rents for the benefit of the treasuryand the consumers (Dnes 1995, Brauetigam1989, Guislain 1997).

Contestable markets: A second way tointroduce competition has been formalized withthe concept of “contestability”. According to thisconcept, if an industry behaves as if it iscontestable (due to the relatively costless entryinto and exit from the industry), most of thebenefits of perfect competition can be attainedwithout government intervention. The essentialidea is that the threat of entry into an industryand potential competition may give anincumbent monopolist effective incentives tobehave as if there were a competitive market.The key aspect of a contestable market and thekey to guaranteeing competitive outcomes istherefore the existence of conditions enablingentry. According to Baumol, Panzar and Willig(1982), if one lowers artificial entry barriers andnew entrants need not incur significant sunkcosts, then all the benefits of competition will beavailable regardless of the market share of theincumbent. The degree of contestability of amarket can then be measured by the share of theinvestment that is composed of sunk capital.Industries with substantial sunk costs such as therailroad industry are therefore not likely to becontestable, whereas industries in which capitalis highly mobile may be contestable. Forexample, in the case of the airlines industry, ithas been argued that airline markets arecontestable since entry and exit is quite easy andthere are virtually no sunk costs in the industry(Braeutigam 1989, Teece 1994, UNCTAD1995).

Intermodal competition: A third way tointroduce competition is through intermodal

competition. For example, in the transportationsector of the economy, monopolistic competitionamong various modes of transport (e.g. railroadsand road transport) is often referred to asintermodal competition. The essential idea is thatif intermodal competition is strong enough, itmay become a basis for deregulation even if oneor more of the modes of transport appears tohave the structure of a natural monopoly. Inrecent years the move toward deregulation of therailroad industry partially results from pervasiveintermodal competition among the railroads andother modes. In other industries similar types ofcompetition have occurred. For example, cableTV, a once heavily regulated industry, haslargely been deregulated, in part because ofheavy competition from over-the-airbroadcasting. The same also applied totelecommunication industry with competitionacross market segments such as mobile and land-based communications. For example, in contrastto the sluggish growth and small size of thestate-owned wireline network, wirelesstechnology has taken off in Africa, fueled largelyby private investors (Braeutigam 1989, Teece1994).

In addition to these alternatives, therealso exists the possibility of introducing“yardstick competition”, which does not obviatethe need for regulation, but facilitates theregulators’ task. Yardstick competition iscreation of entities whose performance can bemeasured, since the performance of one entitycan be compared with that of another. Accordingto this method, a firm with a natural monopoly isbroken up into separate entities which supplydifferent regions. Each entity retains itsmonopoly but only in relation to its own region.Yardstick competition can then be used as aregulatory tool to compare the performance ofthe monopoly operator with that of operators inother regions of the country and withinternational norms; the regulator can use suchcomparative information to justify tougherperformance targets or tariff adjustments at thetime of regulatory review (Ogus 1994, Foster1992).

Would these alternative measures thencompletely obviate the need for regulation?According to Joskow (1998), ‘complete’deregulation policies are not likely to be realisticor effective policy options in most network


industries. In most of the network industriessubject to reform, certain important segmentscontinue to be natural monopolies, and thus‘competition in the market’ cannot be reliedupon to yield satisfactory performance. Inpractice, ‘competition for the market’ throughconcession or franchise contracts must alsoconfront problems resulting from significantsunk costs, asset specificity, and incompletecontracts. Moreover, the effectiveness ofcompetition will depend on policies governingthe initial structure of the competitive segments,the conditions of entry into the market, and theprice and non-price terms and conditions ofaccess to ‘bottleneck’ monopoly networkfacilities for competing suppliers, who need suchaccess to compete effectively.

It is important to note, therefore, thatmarket liberalization is not the same asderegulation (meaning that governments arerelinquishing their regulatory powers).Regulation of natural monopoly industries is stillcrucial. Vogel (1997) finds that there is nological contradiction between more competitionand greater government control.

What is necessary is the redefinition ofthe regulation policy: liberalization requires re-regulation, which implies the reformulation ofold rules and the creation of new ones. In fact,market liberalization and currently on-goingprivatization processes around the worldthemselves bring new regulatory issues to thefore. An example is the need for regulation toaddress the common interconnection problem,for instance in telecommunications. Also inelectricity, as competition moves from thegeneration side to the wholesale or even retailside, issues of third-party access will more andmore come to the fore.

In addition, countries now have tograpple more explicitly with distributionalimpacts, which have to be carefully considered ifthey want to increase the chance of success ofcompetition reform. Public ownership may havebeen selected specifically because it wasconsidered the most appropriate legal form toachieve distributional goals (Ogus 1994). Thisapplies most obviously to utilities, where it maybe felt desirable to supply certain categories ofconsumers at below-cost prices so as to provide auniversal service. Distributional effects resultingfrom “economies of density” following

increasing competition or privatization drivenmainly by efficiency (and budgetary)considerations make it necessary to introducedistributional considerations more explicitly intothe design of regulatory reform, so as tominimize distributional side effects.3

How did developing and transitioneconomies then manage to bring regulatorysystems in line with these new developments andcomplex regulatory demands and to findregulatory approaches that match both theirspecific, new regulatory needs and capabilities?To what extent did these new issues then actuallyreach the political agenda in developing andtransition economies and result in reforms? Thefollowing chapter investigates recent responsesand experiences of developing and transitioneconomies in the area of natural monopolyregulation.

Actual Responses andExperiences inDeveloping andTransition Economies

The actual speed with which the competitivemodel is being advanced in developing andtransition economies has been rather slow,especially when compared with the privatizationprocess itself.

Widespread privatization, butambivalence towards real reform

The recent public finance crises in manycountries, combined with huge investmentrequirements, have made private-sectorparticipation necessary. Furthermore, the poor

3 Economies of density means that costs ofsupplying a particular customer are significantly influencedby the spatial density of surrounding customers. Sincecompetition tends to generate price structures that reflectunderlying costs of supply, one consequence of competitionin network industries is that the prices of physically similarproducts will tend to exhibit quite considerable place-to-placevariations, often creating problems in the pursuit ofdistributional goals (Yarrow 1994).


performance of most public enterprises and theirinability to offer a quality service and meetdemand have encouraged many governments toturn to the private sector for the provision ofinfrastructure services, leading to the need forreforms. However, large companies indeveloping and transition economies that wereprivatized were often sold as monopolies or near-monopolies. Instead of creating greatercompetition in the concerned sectors beforeprivatization, all that has been accomplished issubstitution of a private monopoly for a publicone.

Ideally, privatization of large networkcompanies offers the government a uniqueopportunity to rethink and reform the entireorganization and structure of the sector.Activities or services that were provided by anintegrated, monolithic enterprise will have to beunbundled and competition introduced in thosesegments that can sustain it. Divestiture will veryoften be less important in itself than effectivelydemonopolizing and opening up the sector tocompetition (Guislain 1997). After all, theefficiency impact of privatization depends on thequality of government regulation and its abilityto harness competition for sectoral reform.

While some may be convinced thatprivate ownership leads to greater productivity,many authors such as Stiglitz (1998) find that anenterprise’s efficiency is determined not so muchby its public or private ownership as by theregulatory structure and the degree of ompeti-tion under which it operates. By looking at theexample of China vis-à-vis the former socialisteconomies, he concludes that effectivecompetition and regulatory policies areimportant, rather than privatization itself. Chinahad shown that an economy might achieve moreeffective growth by focusing first oncompetition, leaving privatization until later. Incontrast, competition remains thwarted in manyof the former socialist economies that pursuedprivatization first, demonstrating that withouteffective competition and regulatory policies,private rent-seeking can be every bit as powerful,and perhaps even more distortionary, than publicrent-seeking. Moreover, there are those instancesin which public enterprises have operated at alevel of efficiency comparable to, or greater than,that of similarly situated private enterprises;typically these are associated with firms

subjected to competition, either in exports (as inthe case of Korea’s steel industry) ordomestically (as in Canada’s railroads).

In practice, while privatization oftraditional natural monopolies has becomewidespread in many developing countries overthe past 10 years, their policies towards realsector reform have often been ambivalent.Certainly, there is much privatization, yet theactual degree of commitment to competition-based reform and the measure chosen varyconsiderably among countries and industries.

Some alternative measures to introducegreater competition have taken root in thedeveloping world. Table 1 summarizes differentmodes of privatization and sector reformmeasures in some network industries.

A measure to introduce competition forthe market via competitive bidding ofconcessions for instance, has taken root inpower, telecommunications, railway, and waterenterprises in developing countries as diverse asChina, Guinea, Hungary, and Mexico. Countrieslike Argentina and Chile not only activelyintroduced competition in the market throughvertical disintegration of theirtelecommunication or electric power enterprises,but also adopted yardstick competition measuresin several industries to supplement their sectoralreform efforts. Nonetheless, the breadth, depth,and methods of the private participation as wellas of the sector reform remain highly unevenamong countries and industries. For instance,Argentina and Hungary have chosen to unbundlethe gas sector before privatization, introducinggreater competition, whereas privatization hasnot yet been accompanied by unbundling orgreater competition in the gas industry in Russia.

A sectoral and regional breakdown ofthis highly uneven process of privatization andreform in the developing world concerningnatural monopoly sectors reveals the followingoverall picture.

Telecommunication and electric power leadthe way

The bulk of privatization and demonopolizationhas taken place, above all, in telecom-munications and then in electric power.Electricity has also become a leading networksector in attracting private participation and has


been undergoing increasing restructuring basedon deregulation of key parts of the industry andbreakup of vertically integrated organizationsand systems, through the separation ofgeneration/transmission/distribution. This has forexample already occurred or is planned inArgentina, Chile, Peru, Bangladesh, India, andthe Philippines (Paddon 1998). Private sectorinvolvement in water industry is yet a relativelyrecent phenomenon. Before 1990 privateparticipation in water was rare, except infrancophone countries, and it still remains smallrelative to private participation in other networkindustries (Silva, Tynan and Yilmaz 1998,Izaguirre 1998).

As for the forms of privateparticipation, there is also a significant sectoral

variation. Divestiture of public water and railwayassets is comparatively rare. Few railways havebeen truly privatized. Instead, most governmentshave preferred to concession or franchise theirrailways. In water, concessions are the mostpopular form, where concession contracts haveallowed governments to maintain ownership ofsector assets while delegating substantialresponsibility and risk to the private sector. Mostwater and railway assets remain in the publicsector, and governments are resistant to givingthem up. This highlights the sectoral differencein asset ownership between water and railway onthe one hand and energy on the other (Silva,Tynan and Yilmaz 1998, Thompson and Budin1997).

Table 1: Network industries: Modes of privatization and sector reform(Selected developing and transition economies)

Mode Divestiture Concession andleasing contracts

Introduction ofcompetition in themarket

Yardstickcompetition

Sector (periodic introductionof competition for themarket throughcompetitive bidding)

(e.g., through verticalbreakup of integratedcompanies)

Telecom(wireline voice)Activeprivatization andcompetition-basedreform

Argentina, Chile,Cuba, Guinea,Hungary, Jamaica,Mexico, Peru,Venezuela

China, Cook Islands,Guinea-Bissau,Hungary, Indonesia,Madagascar, Mexico

Chile, Mexico,Philippines

Argentina(basic telephoneservices),Tanzania(basic telephoneservices)

Electric power(generation)Active privateparticipation andunbundling

Argentina, Bolivia,Chile, Hungary,Pakistan, Peru

China, Ivory Coast,Guinea,Hungary, Mexico

Argentina,Bangladesh, Bolivia,Chile,India, Peru,The Philippines

Argentina(distribution)Chile(distribution)

Gas(transport anddistribution)

Hungary, Latvia,Russia

Argentina Argentina,Hungary

Argentina(distribution)

RailwaysMainly franchising

Bolivia Argentina, Brazil,Ivory Coast-BurkinaFaso, Chile, Mexico

Water(distribution)Relatively smallprivateparticipation;concessionpreferred

Argentina, Brazil,Chile, China,Colombia, IvoryCoast, Guinea,Hungary, Macao,Malaysia,Mexico, Senegal

Note: The table includes only countries that have privatized by transferring existing public-sector facilities to the private sector, not those thathave opened up the sector in question through greenfield concessions or BOT and BOO contracts only, such as Thailand(telecommunications) and China (power generation).

Source: Dnes (1995), Guislain (1997), Nells and Roger (1994), Otobo (1998), Paddon (1998), Plane (1998), Thompson and Budin (1997).


Latin America and East Asia dominate

In most network industries (such as energy,water, and telecommunication), Latin Americaand East Asia (including the Philippines andMalaysia) dominate private sector participationtrends and restructuring process; and within eachregion, a few countries lead the way (Silva,Tynan and Yilmaz 1998).1) In particular, the Latin American region has

a rich fund of experience in privatizationand restructuring of natural monopolyindustries. A few leading countries, such asArgentina, Chile, and Peru, privatized majornetwork industries relatively early on as partof broader economic reform programmes, inorder to overcome major bottlenecks caused

by the inadequacy and poor state of publicutilities (Guislain 1997). Among LatinAmerican countries, Argentina is thecountry that has gone furthest in matters ofprivatization in Latin America since 1990and has been at the forefront of variousnetwork industries’ reform process, byintroducing competition in the marketthrough vertical disintegration.

2) In Asia, despite its clear dominance ofinvestments in projects with privateparticipation, the divestiture trend has notbeen as pronounced. Few countries therehave adopted or implemented large full-scale privatization programmes. Even intelecommunications, divestiture of thedominant operator has been partial with the

BOX 1: Importance of ensuring real competition:Comparative experiences of Chile and Argentina (electric power)

Argentina is the country that has gone furthest in introducing full competition and vertical disintegration inthe electric power industry. Chile is also a path-breaker of privatization in the developing world—alongsideArgentina. Yet in the electricity sector, the restructuring of enterprises prior to privatization fell short ofwhat was needed to ensure competition. Despite the comparatively advanced Chilean regulatory framework,it could have paid more attention to the property structure, to ensure real competition.

• Chile: In the case of electricity, Chile was committed to vertical disintegration, but to a lesser extent tocompetition. In Chile, there were no restrictions on cross-ownership of assets in different segments,unlike Argentina (and Peru), which has prohibited any company or group from controlling more thanone of the market segments (e.g., electricity generation, transmission, and distribution). One investmentgroup controls most of the system’s generating capacity, the largest distribution company, and thetransmission assets. Cross-ownership and consequent conflicts of interest have hindered thedevelopment of a more competitive generation market.

• Argentina: In Argentina, the power sector was restructured radically in 1992 by unbundling generating,transmission, and distribution activities and organizing them under separate companies. Joskow (1998)describes Argentina’s approach to electric power as a “big bang-approach”, in which privatization,restructuring, and the introduction of competition were all accomplished in one big step.Argentina, privatizing much of its power system more than ten years after Chile, benefited greatly fromobserving that country’s problems associated in particular with cross-ownership. Argentina separatedmonopoly transmission and distribution segments from the competitive generation segment. It adopteda mandatory separation principle. No generator is permitted to control more than 10 per cent of thesystem’s capacity, and restrictions on reintegration and cross-ownership are enforced. The resultingdiversity in ownership ensured a more competitive environment for generation than in Chile. Therestructuring programme in Argentina created a large number of private generating companies, andcompetition at the generation level has been intense. Transmission and distribution became regulatedprivate monopolies. Retail tariffs are regulated through a price cap mechanism (essentially RPI-X,where RPI is the retail price index and X is productivity gains, with X adjusted after five years). TheArgentine privatization has been a clear success in electricity industry.

Source: Bitra and Serra (1994), Chisari, Estache, and Romero (1997), Guislain (1997), Joskow (1998), Labor and Garcia (1996).


government continuing to be the controllingshareholder (e.g., Telekom Malaysia with apublic floatation of 25 per cent of the sharesin 1990) (Guislain 1997). China has alsotaken a cautious approach to opening certainsectors such as electricity to privateinvestment, and mainly relied on jointventures between private sponsors and state-owned enterprises (Izaguirre 1998). Themajor rationale given for network industryprivatization in the region has been the needto introduce the additional resourcesnecessary to extend access to the service,improve service quality and modernize thesystem (Paddon 1998). There is littlesubstantive evidence in the region of animprovement in the quality of utilityservices after privatization, leading toquestions regarding the guarantees foradequate service quality and pricing writteninto privatization arrangements.4

3) In Africa, a number of supplements toprivatization have been explored. Inparticular, leasing contracts and concessionsare viewed as promising arrangements,which provide an inducement in that theyplace an appreciable part of the risk on theprivate operator. Some African countriessuch as Ivory Coast already provide someexamples of these forms of privatization.African privatization has been most of allmotivated by the new financial constraints.Insufficient public money is forthcoming,putting African governments in difficulty infinancing the development of utilities, withthe result that African governments areunable to improve their public services. Forelectricity alone, according to the WorldBank, those governments will need to investa total of US$17 billion between now andthe year 2005. African States themselves

4 In Manila’s water privatization, NGOs claim thatthe contracts with concessionaires provide for inadequatehealth standards, and lack appropriate environmentalstandards. In Pakistan, there are complaints about electricityprivatization resulting in doubling of electricity prices in oneyear so as to accommodate demands by foreign investors forhigher profits paid in foreign currencies. There are also theconcerns about the pricing agreements reached by thePakistan Government with transnational corporations toinduce them to invest in new independent power producers.As far as the effects on price are concerned, in each of theutilities, the evidence from across the region is thatprivatization and restructuring are associated with increasesin prices and charges for some consumers and have generallybeen disadvantageous to domestic consumers (Paddon 1998).

will not be able to provide more than US$5billion and funding sources US$2 billion(Plane 1998).

4) In transition economies, privatization ingeneral has been a tool of transition. It hasbeen used to establish property rights, toform a private sector and the basis of amarket economy, to enable efficientgovernance and management of formerlystate-owned enterprises. Yet, privatization ofnatural monopoly sectors was usually notfeatured during the early years of reforms.Instead, reduction of price subsidies hasbeen a feature of transition in somecountries (e.g., Lithuania), partly inpreparation for privatization. In particular,increasing energy prices to cover costs, andincrease profits, has been a painful processin many transition economies. The generaltrends with regards to energy privatizationfor transition economies are to movetowards increasing prices; decentralizingdistribution to local authorities; and someprivatization, especially production. Privateparticipation in electricity has beenconcentrated in the Czech Republic,Hungary, Kazakhstan, and the RussianFederation with some vertical unbundling ofexisting firms. Privatization of water in theregion has so far been restricted largely totwo countries, the Czech Republic andHungary, with a couple of cases in Poland.Restructuring by decentralization has takenplace more extensively, though thisdecentralization has probably reducedefficiency. In particular, the problemsencountered by competitive restructuringinitiatives in Russia point out the followingbarriers to reforming natural monopoliesthat are particularly important in Russia and,by extension other transition economies: thefirst is political opposition from themanagement of the firms themselves (e.g.,Gazprom). The second obstacle to reformednatural monopoly regulation lies with thesubnational authorities. The regionalauthorities’ dual role as owners of regulatedfirms and as the principals to which theregional regulatory commissions aresubordinated has not worked well.Moreover, much of the so-called“privatization” has really been the transfer


of ownership rights from the federal toregional governments. The problem is thatsuch transfers have introduced additionalelements of confusion into corporategovernance, and created conflictingincentives for federal and regional agenciesthat function both as owners and asregulators. This confusion and theconflicting incentives have been a majorobstacle to regulatory reform in Russia’snatural monopoly sectors (Izaguirre 1998,Martin 1997, Slay and Capelik 1998).

How to better regulate naturalmonopolies: Highlighting some policylessons and regulatory experiences

The efficiency and behaviour of a monopolisticenterprise, whether private or public, dependsmuch on the framework in which it operates, andespecially on the existence of performance-enhancing incentives and penalties (Guislain1997). We should acknowledge that neither thesuperior performance of a public monopoly to aprivate monopoly nor the contrary has ever beenproven empirically.

Which specific approaches andtechniques need to be applied? This is the crux ofthe matter of designing natural monopolyregulation policy. In this section, we willhighlight and analyze some regulatory practicesand experiences of developing and transitioneconomies, so as to draw some policy lessons.

Harnessing competition for regulation asa goal

We observe from the experience of manydeveloping countries that competition is anefficient form of regulation. Where privatizationhas gone with strong competition in the market,the outcomes were positive, as is the case withArgentine electricity (see Box 1). Thus,whenever possible, harnessing competition in themarket for regulation should be the main goal.

Try alternatively “competition for themarket”

If competition in the market is not possible, ase.g., in the water industry, one should at leastorganize the sector so that it can take advantageof opportunities for competitive bidding.

In the water industry, network-relatedcosts are a higher proportion of total costs thanin gas, electricity, or telecommunications, andthe gains to be made from introducingcompetition by splitting up ownership of thesystem are relatively small. Thus, most waterwill be supplied monopolistically at least for thetime being, and franchising appears as a way ofencouraging efficiency despite the monopoly(Klein and Irwin 1996). Argentina’s positiveexperience with an international competitivebidding process for Buenos Aires waterconcession in 1993 is a case in point. InArgentina, water and sanitation competition hasbeen introduced through a bidding process,closely resembling Demsetz-competition(Chisari, Estache, and Romero 1997).

In the railway industry, franchising isalso a preferred practice. The success of the earlyconcessions and the lack of credible alternativeshave caused a snowballing of such franchise-based reforms in Latin America, spreading alsoto other regions. So far the experiences inArgentina, Brazil, Chile, Mexico, and IvoryCoast-Burkina Faso are encouraging (see Table 1again) (Thompson and Budin 1997).

Yet franchising is no panacea.Introducing competition for the market requirescareful supplementary regulation.

First of all, it requires a substantialgovernment investment in the initial design ofthe concession. This also entails government’sfundamental decision concerning the degree offlexibility of the concession agreements to beallowed (see Box 2 for some guidelines).Governments still have to deal with the familiarproblem of price regulation. At the time of theconcession, the regulator must try to estimate theright price e.g., for water (see the followingsection 2.3).

In addition, over the course of theconcession, it inherently requires continuinggovernment involvement in regulating safety,monopolistic behaviour, and compliance with thepricing and service requirements of theconcession. It cannot simply walk away from itsconcessions once they are completed (Thompsonand Budin 1997).


Regulating monopoly price: Cost-basedor price-based formula?

Setting the optimal price for natural monopoliese.g., at the time of the concession is not an easymatter. Difficulty with monopoly pricing resultsfirst of all from the problem of regulators nothaving access to good information, regardingdemand and best practice cost conditions(Bradburd 1992). Secondly, there is difficulty indesigning a system of price controls that gives astrong incentive for the regulated firm to investmore and to improve its efficiency.5

5 In particular, the main purpose of tariff regulationin most developing countries should be to foster investmentrather than to control the level of prices per se. In these

Three basic issues are involved:1) the rate level issue—making sure that the

total earnings of the firm are appropriatelyrelated to the costs;

2) the rate structure issue—the determination ofthe proportion of earnings between differentservices and different customers;

3) the quality issue, to ensure that price controlsdo not create incentives for firms to reducethe quality of products and services.

countries it may indeed be preferable to have relatively hightariffs. An enterprise could then self finance a large part of itsinvestment programme, and contractual or regulatorymechanisms could compel it to reinvest the “excess” tariff inthe sector to meet demand (Guislain 1997).

BOX 2: Designing the initial concession agreements—flexible or inflexible?(Lessons from Guinea, Ivory Coast, Peru, and Venezuela)

It is not easy to find a balance for each country and each sector between restrictive rules and adoption of amore flexible framework that allows for evolution of the rules but adds uncertainty.

Generally speaking, detailed a priori regulation is better suited to relatively stable, technologicallymature, and monopolistic sectors, such as water, than to sectors undergoing rapid technological evolution,such as telecommunications.

However, in developing countries with weak administrative and judicial systems or poor trackrecords concerning credibility, the use of detailed and relatively inflexible concession agreements with fairlyprecise upfront regulation may be preferable to more flexible rules subject to more discretion on the part ofthe regulator. This may be more likely to reassure investors than the creation of an autonomous regulatoryagency with discretionary rulemaking powers.• Guinea and Ivory Coast both opted for the inflexible approach in privatizing their water supply and

electric power sectors; the leasing contract and concession agreement were accompanied by a detailedschedule of obligations and conditions, leaving few aspects to be decided or agreed upon duringexecution of the contract. The results are so far encouraging.It may be desirable to anchor the regulatory framework securely in a law, which would give it a greatstability, though little flexibility, as Peru did. Peru needed to establish a reputation for credibleregulatory rules to attract investment to the sector. The terms and conditions of the initial regulatorycontract are enforceable under commercial law, giving the regulator little discretion during theexclusivity period. It has been successful by and large, exceeding all of its major investment and serviceimprovement goals. The regulatory framework, including the terms and conditions of concessioningcan be spelled out e.g., as a sector-specific privatization law. This can be particularly useful forgovernments with low credibility and an inadequate track record, which will usually have to offer moreguarantees to attract private investors.

• In contrast, the lack of such institutional and legal anchoring probably remains one of the majorweaknesses of Venezuela’s telecommunication franchising. In 1994 relations between CANTVtelephone company holding a 35-year concession, on the one hand, and the regulator and government,on the other, became very tense. For political reasons, the regulator blocked the rebalancing of rates anddid not meet deadlines to authorize some rate increases provided for in the privatization agreements;one of the quarterly increases was even denied. Even if the short-term effect is not clear, it is likely thatthis interference will be detrimental to continued private investment.

Source: Joskow (1998), Guislain (1997), Plane (1998)


Of the various methods of monopolypricing regulation, those applied often toconcessions in developing countries aredescribed below.

Rate of return (or cost-based) regulation

This “cost-plus-fair rate of return” regulationmethod (based on average cost pricing) allowstariffs to rise subject to a predetermined rate ofreturn. Prices are adjusted so as to keep thecompany’s rate of return on capital at a constantlevel. If the company’s rate of return falls belowthat level, the regulator allows prices to rise. Thiswas popularized in the United States and hasbeen transferred to several developing countries.The telecommunications sector in the Philippinesis one example. Under this approach, the tariff iscalculated so as to cover the regulated firms’operating costs, plus a rate of return on theinvestment.

The problems with this price regulationmethod are that:

One must estimate cost of buildingcapacity. The basis of the calculation may beinflated by means of unrealistic or spurious costsor investments. Under this approach, the utilitycalculates—and the regulator reviews—theexpected operating cost for a normal year(information problem).

It is considered to provide very littleincentive for the regulated firm to reduce costsand improve technology. This does notencourage firms to minimize cost, either. It alsooften encourages to overinvest in capital(incentive power problem).

It is complicated to administer. Itrequires extensive research into an enterprise'saccounts--and thus plentiful human resources--todetermine which costs should be included in therate base, and which should be disallowed. Itrequires constant monitoring of management andcontinual negotiation between the two sides(regulation cost problem).

Its tendency to distort input choices, aswell as its administrative difficulties, have madethis method of regulation increasingly unpopular.In particular, in most developing countries whereprofessional skills are scarce, the opportunity

cost of scarce human capital devoted toregulation is too high to recommend its use(Jones 1994, Klein and Irwin 1996, Yarrow1994).

Price cap (or price-based) regulation

This regulation method tries to avoid theseproblems. It emerged during the privatization ofBritain’s utility industries in the United Kingdomin the mid-1980s and is now used in developingpublic utilities in countries such as Argentina,Brazil (new law on concessions), and Chile.

Under a price cap, prices are allowed torise by means of a formula, known as RPI-X,that increases the tariff by the increase in theretail price index adjusted by an efficiencyfactor, X, to account for expected productivitygains and other changes. Under this method, thecompany has an incentive to lower costs, since itkeeps the resulting profits, because it allows thefirm to hold on, at least for a designated period,to the profit gains from cost reductions.

The aim of this system is basically togive the regulated firm an automatic incentive toimprove productivity, while at the same timeenabling consumers to benefit from suchimprovements through the tariff cuts introducedat times of revision.

Fixing the initial tariff is accompaniedby an automatic adjustment rule valid for a givennumber of years. During an initial period, theadvantage of all such gains goes entirely to theconcessionaire. In return, any real cost increasesare not passed on to the consumer, except inunusual circumstances such as sharply higherpurchase prices for energy.

The flexibility and the relatively greaterease of administration have made price caps apreferred form of regulation for governments.Price caps allow a company to adjust pricesquickly when market or competitive conditionsrequire it, because an extensive review of costsand earnings is not required. Instead, price capprovisions enable a utility to adjust prices as itwishes, provided the average price for aspecified basket of services does not exceedsome maximum value (Klein and Irwin 1996,Warrell 1997).


RPI-X price adjustments certainly seemfar superior to rate-of-return price adjustments,but the real difference between them is not as bigas it might seem (Ergas 1994, Jones 1994,Yarrow 1994). For example, according to Jones(1994), the incentive power of a monopoly-pricing scheme to induce efficiency does notdepend on whether it is couched in cost-plus orRPI-X terms. Rather, it depends on the length ofthe regulatory lag and the expectations of howprices will be adjusted at the end of the lag. Theregulatory lag is the period for which the price isset. The longer the lag, the higher the incentivepower of the system. One advantage of RPI-X isthen that it is typically associated with arelatively long lag. However, in practice, thepoint is that this dimension of choice can easilybe added to cost-plus schemes. Allowing theprice to vary with the rate of inflation promotes a

longer lag before prices have to be established.But again, the adjustment can also beaccomplished by choosing a price index thatrelates more specifically to the input priceinflation experienced by companies, as seen inChile (see Box 3).

This price regulation method raisescomplex issues about the level of the cap, theservices to be covered and the monitoring ofservice quality; hence also imposing a heavy costin supplementary regulation. For example, RPI-X formulas need to be reviewed every three tofive years or so, since the regulator does notknow exactly how large X should be and, inreviewing whether X was set appropriately, willtake into account the profits being made by thefirm.

There is additional trade-off confrontingthe regulator. Price cap regulation may be good

BOX 3: The two boundaries of possible spectrum of pricing techniques:New Zealand (minimum) and Chile (maximum)

According to Jones (1994), New Zealand’s effort represents the minimum that should be done, while Chile’sis the maximum that should be attempted. Most countries fall somewhere in between, with a relatively longregulatory lag and X set using as much exogenous data as possible. The precise point on that continuumwould then be a function of specific country and industry conditions.

1) New Zealand (model of “simplicity”): The system used in New Zealand (described as “regulationwithout regulators”) is extremely simple. It is based on the RPI-0 pricing mechanism, which isextremely economical in terms of the cost of regulation. Its cost-incentive power is quite high becauseof the indefinitely long regulatory lag with a fixed price in relation to inflation. The primary emphasis ison cost-efficiency incentives, with considerably weaker controls on the allocative inefficiencies ofmonopoly pricing. It minimizes the costs of the major regulatory failures, though at the expense ofallocative form of market failure.

2) Chile (model of “sophisticated specificity” plus yardstick pricing): The system can be described as cost-plus-fair-return because firms are allowed a rate of return equal to the risk-free rate plus a premiumbased on the systematic risk of the industry and the difference between the risk-free rate and the returnon a diversified investment portfolio. (It could as well be described as RPI-X because the price cap isadjusted every two months to reflect inflation). Which phrase is chosen is immaterial because the realdistinction lies elsewhere. The adjustment period is explicit and reflects a long lag of five years.Moreover, a range of regulatory technology is spelled out in law. In addition to the sophisticated fair-rate-of-return and inflation-adjustment mechanisms, long-run marginal costs are calculated in thecontext of a five-year investment plan (designed to minimize the system costs of meeting projecteddemand), markups from marginal to average costs are apportioned via Ramsey pricing, and so on. Inaddition, Chile used yardstick competition as the local best-practice benchmark as a complementarymeasure (e.g., in the case of electricity distribution). If the regulator and the firm disagree, they can alsoappeal to a technical arbitration board. Chile’s system has resolved the market failure problem, but at anincreased cost of regulation and at the expense of complexity and information requirements.

Source: Jones (1994)


for cost reduction incentives, but may be bad forquality incentives. Incentives for cost-reductionmay translate into a tendency to chisel onquality, then leading regulators toward greaterinvolvement in investment and product qualitydecisions, if not so much concerning the detailsof tariff formulation.

The challenge for the regulator is thenhow to balance and reconcile all these differentproblems such as information requirement andheavy cost of regulation. The choice of adequatepricing technique is complex and there is no bestcase for all circumstances.

The feasible choice for pricingregulation design lies on the continuum betweenNew Zealand and Chilean pricing systems,which define the boundaries of what might bepractical. New Zealand’s model of extremesimplicity and Chile’s model of sophisticatedspecificity mark the end points of the spectrumof the currently feasible pricing models. Manyintermediate solutions are possible between thesetwo models (see Box 3).

A complementary approach is then touse a benchmark or yardstick against which the

enterprise measures itself, thus reducinginformation requirements from the regulatedfirm. Using yardstick competition can alsoreduce the undesirable incentive effects of bothRPI-X and rate-of-return pricing models (Yarrow1994, Klein and Irwin 1996).

Getting the relation between privatizationand regulation right: A few guidelines

In addition to all these measures of introducingcompetition, it is equally important that one getsright the privatization sequence and coordinationwith regulatory reform.

Here are a few guidelines based on theobservation of countries’ experiences.1) Implementing related structural andregulatory reforms upfront and prior toprivatization is important. The regulatoryframework should be as little ambiguous aspossible and must be completed prior toprivatization.Regulatory reform and privatization processesneed to be closely coordinated, and theirsequencing and coordination will have to bethought through from the outset (Bitrain and

BOX 4: Using yardstick competition as a “complementary” measure(Examples: Argentina, Chile, Tanzania)

Argentina used this technique in several sectors. In addition to introducing competition in the market whereit was deemed feasible, Argentines decided also to break up existing monopolies on a geographic basis tocreate benchmark (or “yardstick”) competition in most infrastructure sectors. In the telecommunicationssector, for instance, it was decided that direct competition should not be introduced immediately for basictelephone services previously provided by ENTEL. ENTEL was split between two geographic areas (northand south, with Buenos Aires divided into two zones), served by two separate privatized companies.Although direct competition is (initially) not authorized for basic services, this geographic division allowsthe regulator and the public to compare the performance of the two companies and exert pressure on the lessefficient operator. The same principle was applied to power and gas distribution companies (Guislain 1997,pp. 214-15).

Another example is the power sector in Chile, where regulators have devised a pricing structurebased on the cost structure of an "optimized" distribution firm. Distributors measure their costs against thoseof the model firm. This method is therefore particularly useful for encouraging efficiency (Nells and Roger1994, pp. 10-11).

Tanzania also provides a good example of horizontal unbundling based on geographic location, forcellular services. The regulator has divided the country into four zones and allowed service providers ineach. Millicom (Tanzania) Ltd. is licensed to provide service in Dar es Salaam and Zanzibar. TRITelecommunications Tanzania Ltd. is to provide the coastal (Dar es Salaam) and Northern Zones. TanzaniaTelecommunications Company Ltd. is to provide services in Northern, Central and Southern HighlandZones; and MIC Tanzania is providing mobile cellular telecommunication services in the coastal area of thecountry. This method allows competition by comparison, by forcing each of the regional operators to revealmuch data on key areas of their operations (Otobo 1998, pp. 24-25).


Serra 1994, Guislain 1997). The privatization ofthe Argentine telecommunications operator,ENTEL, in 1990 provides a good example of theimportance of establishing a regulatoryframework before privatization proper. Partiallydue to a conscious decision on the part of theArgentine Government to give priority to aspeedy conclusion of the sale, the regulatoryregime was not defined until the very end of thebidding process for ENTEL, following severalmajor modifications during the process itself.This regulatory failure had a negative impact onthe telecommunications sector, as shown by theproblems with the revision of the tariff formula(UNCTAD 1995, p. 136). There are also otherexamples such as water privatization in Manilain the Philippines (see Paddon’s case study 1998,pp. 77-79), which shows the establishment of aclear and effective regulatory framework as apre-requisite for the success of a naturalmonopoly privatization programme.2) The emphasis and priority should thusbe on the competition-based reform of the sector,rather than on the transactional aspects of thedivestiture of one or more individual publicenterprises. Many privatization programmesappear to focus more on revenue generation thanon the longer-term gains that more radicalrestructuring of the enterprise or sectorconcerned would bring (Guislain 1997).According to Paddon’s study (1998), theprivatization practice in Asian public utilitieshitherto is generally dominated by gross figuresof the overall transactions. In most cases, there islittle evidence of specific assessments of a widerange of potential costs or benefits of theprivatization of the utility3) Defining privatization objectives is animportant exercise that should be undertaken asearly as possible. This is particularly necessary,given the multiplicity and sometimes mutuallyincompatible nature of the objectives.Anunderstanding of the possible conflicts betweenallocative efficiency and other objectives isessential. For example, the sales proceeds to thegovernment may be enhanced by selling a largeenterprise as a single entity, whereasrestructuring the enterprise into smaller units willimprove the competitiveness of the sector andthe economy but reduce the proceeds of the sale.Many privatization programmes have flounderedwhen clear objectives were lacking or where

conflicting objectives were simultaneouslypursued (Guislain 1997, Bradburd 1992,Waddell 1997).

Dealing with the problem of regulatorycapacity: Privatization of regulatory tasksas a solution?

Regulatory capacity is an essential prerequisitefor managing privatization and implementingcompetitive restructuring of natural monopolies.

However, history teaches us that fewcountries have had the essential capacity tohandle a complex regulation policy. Historically,the emergence of municipal ownership of theutilities was due, in part at least, to the lack ofconfidence in the early regulatory bodies; andone reason for the nationalization of the railwayswas the failure of the pre-war regulatory systemto meet the needs of the industry. More recently,even the relatively sophisticated regulatoryagencies, established to control the prices set andthe profits earned by the industries privatizedunder the Thatcherite programme in the UnitedKingdom, have been beset by difficulties, andtheir decisions have been widely criticized (Ogus1994).

There are plenty of examples of thefundamental problem with lack of regulatorycapacity, which hampered privatization andreform initiatives (e.g., the privatization of SriLanka’s bus transport, Wanasinghe andWanasinghe 1991).

Indeed, setting up and choosing theappropriate institutional framework of regulationpresents a major challenge for developing andtransition economies.

For example, a choice has to be madebetween multisectoral regulatory agencies (as inJamaica and Malaysia) and single-sectorregulatory agencies (as in Argentina), whichcreated a regulatory agency for each industry.Most experts agree that a multisectoral agencyoffers advantages over the alternatives. It poolsscarce regulatory resources such as regulatoryeconomists and lawyers, especially important incountries with limited regulatory capacity. Also,by pitting interest groups against one another, itobviously tends to increase resistance toregulatory capture and political interference andfacilitate a more harmonized approach indifferent sectors (Estache 1997).


Even so, whether or not a countryshould adopt one or the other model ofregulation should be based on a number ofconsiderations such as the number of operatorsin the sector; the size of market, the availabilityof regulatory resources, the complexity ofregulatory rules to be monitored and enforced,and the political disposition regarding degree ofautonomy for the regulatory agency. This meansthat the choice of regulatory institutionalframework to be adopted should be guided by itsgood fit with the national context. In general, thelarger the economy, the greater the number ofoperators in the sector or the more complex theregulatory rules for a sector, the greater the needfor an independent sector-specific regulatoryagency (Otobo 1998).

Additional issues facing governmentsconcern the form of the regulatory body, fundingand legal authority, in particular in connectionwith the important issue of ensuring effectiveindependent regulation.

The importance of independentregulation cannot be stressed too strongly. Theexperience of the Hungarian electricity sector—where regulatory agencies in 1996 reneged onpreprivatization promises guaranteeingforeigners an 8 per cent real dollar return oninvestments made in 1995—is instructive in thisregard (Slay and Capelik 1998). Evengovernments with an ambitious privatizationinitiative like Malaysia could not resist politicalinterference (see Box 5).

An independent regulatory authority iscertainly the most attractive solution forinvestors, as it offers a more stable environmentfor privatized natural monopoly firms. However,it may not be applicable to all countries,particularly in many developing and transitioneconomies where there is no tradition ofindependent institutions, free from politicalinterference. It is particularly important that theauthority be granted an independent source offunding. Independence also requires that, wherestate-owned enterprises are operating in thesector, the regulatory function be clearlyseparated from the exercise by the government ofits ownership functions. Where sectors were runas public monopolies the confusion fromcombining operating and regulatory powers in asingle entity or person was not uncommon; as asector starts to open up to new firms, however,this situation quickly becomes untenable(Guislain 1997, UNCTAD 1995).

According to Guislain (1997), theproblem however lies in the reality of manydeveloping and transition economies. In manycountries it is difficult to achieve thatindependence in practice, at least in the shortrun.

The concept of regulatory bodiesindependent of the industry and the governmentis undoubtedly attractive, but independent,autonomous regulatory agencies with decision-making powers may not be suitable for allcountries. If the political independence of the

BOX 5: Importance of independent regulation(The case of Malaysia)

Malaysia’s ambitious and wide-ranging privatization programme covering railways, the national airlines,telecommunications, electricity and water services attracted considerable domestic and foreign investment.This also allowed the Government to shift to the private sector the considerable cost of improving thecountry’s infrastructure needed to sustain its high economic growth. However, questions have been raisedrelating to the role of the Government in promoting the healthy development of the privatized utilities. InMay 1995, the Government—which retains a 70 per cent share in the privatized electricity utility TenagaNational—decided not to allow Tenaga to raise its prices, thereby contravening a 1993 agreement allowingthe utility to adjust its charges according to movements in fuel prices and other costs. The decision wasinfluenced by the Government’s concern that higher electricity prices would add to the inflationarypressures in Malaysia’s economy. On the other hand, Tenaga defended its proposed price increase on thegrounds that an increase in revenues was needed for a multi-million dollar modernization scheme. Manyinvestors now feel that the decision has set an unhealthy precedent for future government interference in theprivatized industries. This case illustrates the importance of effective independent regulation (UNCTAD1995).


regulatory organ cannot be ensured (e.g., as incountries with authoritarian governments),creating a new agency with decision-makingpowers may needlessly complicate themanagement of the sector by introducing anadditional actor and yet another level ofuncertainty. For instance, the United Kingdommodel, and in particular the decision-makingpowers given to individual and independentregulators, should be seen in the proper context:the United Kingdom is a sophisticated industrialcountry with very strong, well-established legalpractices and traditions. In this regard, Guislainrecommends for some developing and transitioneconomies with little or no regulatory trackrecord the adoption of a light-handed system ofregulation with limited discretionary powers andthe contracting out of much of the regulatorycontrol and verification work to reputable privateauditors.

In particular, the idea of a small, centralgovernment team that contracts out importantregulatory tasks to external auditors andinstitutions (i.e., “privatization of theprivatization and regulation process”) clearly isan interesting option and deserves someattention. Complex regulatory functions need tobe performed professionally; where limitedadministrative capacity (as seen in theprivatization of Sri Lanka’s bus transport) isindeed a binding constraint, at least in the shortand medium term, “privatization of regulatorytasks” should be considered. While creation of aseparate group or agency with extensive powersand a clear mandate seems to be the bestsolution, at least for countries with extensiveprivatization programmes, this option will oftenbe better suited to some other developing andtransition economies’ administrative capacity.

Conclusions andRecommendations

Natural monopoly sector privatization is arapidly evolving field, and it would be prematureto venture definitive conclusions as to the “bestpractice” privatization and regulation models fornatural monopolies. Yet, the regulatoryexperiences in developing and transition

economies to date suggest the followingpreliminary guidelines for natural monopolyregulation policy:

The bottom line is that regulation is acontinuing process, whichever model is used.Harnessing competition for regulation should bethe goal, but even the alternative measures ofintroducing competition (such as Demsetz-competition) require substantial supplementaryregulation efforts of government. Embracing thecompetition principle as much as possible isimportant. Yet, liberalization requires re-regulation (i.e., reformulation of old rules andthe creation of new ones). Again, marketliberalization is not the same as “deregulation”.

One should try to take full advantage ofthe unique opportunity that privatization offersthe government to rethink and reform the sector.This implies that one needs to focus more on“reform” based on “real” competition than justthe “private” or “public” ownership issue (Whereprivatization has involved competitive markets,the outcomes have been positive. As far as thepublic vs. private ownership debate is concerned,the superiority of neither ownership structure fornatural monopolies has ever been provenempirically).

To this end, countries must find whichsegments of an industry have competitivecharacteristics and determine the most suitableways of introducing more competition, whilestill maintaining appropriate oversight (e.g., forthose sectors that exhibit natural monopolycharacteristics (For example, Demsetz-competition in combination with yardstickcompetition in the water industry seems so far afeasible way of introducing competition, andvertical unbundling of electricity, hencecompetition in the market seems both feasibleand efficient). This entails that governments stillneed to deal with the thorny problem ofmonopoly pricing.

The price-based (price cap) regulationmodel appears far superior to the cost-based(rate-of-return) regulation model, e.g., in termsof its incentive power. However, the realdifference between them is not as big as it mightseem. In practice, the difference becomes oftendiluted. The feasible choice for pricingregulation design lies rather on the continuumbetween New Zealand’s model of extremesimplicity and Chile’s model of sophisticated


specificity. Many intermediate solutions arepossible between these two models, and theprecise point on that continuum would then be afunction of specific country and industryconditions.

Countries now also have to grapplemore explicitly with distributional impacts, so asto increase the chance of success of competitionreform. While it is possible, and perfectlylegitimate, to argue that the losses made by somegroups are outweighed by the overall gains(according to the so-called “compensationprinciple”), the distributional consequences ofcompetitive restructuring processes (e.g., as aresult of “economies of density”) then need to bemade explicit and at least discussed. Certaindistributional inequities are better dealt with bymeans of subsidies from the government budget.

Not least in this regard, anunderstanding of the possible conflicts betweenallocative efficiency and other objectives(including distributional equity objective) isessential. Given the multiplicity and sometimesmutually incompatible nature of the objectives,particularly between static efficiency anddynamic efficiency, clear definition of objectivesfrom the very outset is important. In case ofeventual trade-off between static and dynamicefficiency (e.g., when determining the right pricelevel), it is important to keep in mind that theultimate objective of natural monopoly industriesis sufficient and sustainable provision of theirservices. While reforms may be necessary tomake services more efficient and economical(hence increasing static efficiency), the priorityshould still be dynamic efficiency.

In addition to introducing greatercompetition, it is also equally important that onegets the privatization process right. In manycountries, privatization seems an unavoidable

outcome of constraints, particularly financialones. Once decided for privatization, propersequencing of the privatization and its coordina-tion with the regulatory reforms are important.The best procedure is that structural andregulatory reforms are implemented upfront andprior to privatization.

It is not easy to find a balance for eachcountry and each sector between restrictive rulesand adoption of a more flexible framework thatallows for evolution of the rule but adds un-certainty. Generally speaking, detailed a prioriregulation is better suited to relatively stable,technologically mature, and monopolisticsectors, such as water, than to sectors undergoingrapid technological evolution, such as telecom-munications. However, in developing countrieswith weak administrative and judicial systems orpoor track records concerning credibility, the useof detailed and relatively inflexible concessionagreements with fairly precise upfront regulationmay be preferable to more flexible rules subjectto more discretion on the part of the regulator.This may be more likely to reassure investorsthan the creation of an autonomous regulatoryagency with discretionary rulemaking powers.

Essentially, there is limited experiencewith regulation and privatization in developingcountries, and we still have much to learn. Thechoice of regulatory approach is complex andthere is no best case for all circumstances. Thus,the choice of particular regulatory framework(e.g., between multisectoral and single-sectoragencies) to be adopted should be guided by its“good fit”, particularly with the particularnational context, and reflect the “reality” ofdeveloping and transition economies. In thisregard, the option of “privatization of theregulatory task” certainly cannot be dismissedout of hand.


ReferencesBaumol, W.J., Panzar, J.C. and Willig, R.D. (1982),

Contestable Markets and the Theory of IndustryStructure, New York: Harcourt BraceJovanovitch.

Becker, G. (1997), There’s nothing natural about “natural”monopolies, in: Business Week, October 6, 1997.

Bitra, E. and Serra, P. (1994), Regulatory issues in theprivatization of public utilities. The Chileanexperience, in: The Quarterly Review ofEconomics and Finance, Vol. 34, Special Issue,Summer 1994, pp. 179-197.

Bradburd, R. (1992), Privatization of Natural MonopolyPublic Enterprises. The Regulation Issue, PolicyResearch Working Papers WPS 864, The WorldBank.

Brauetigam, R. (1989), Optimal policies for naturalmonopolies, in: R. Schmalensee and R.D. Willig(eds.), Handbook of Industrial Organization,Volume II, pp. 1290-1346.

Chang, H.-J. (1997), Critical survey: The economics andpolitics of regulation, in: Cambridge Journal ofEconomics 1997, Vol. 21, pp. 703-728.

Chisari, O., Estache, A. and Romero, C. (1997), Thedistribution of gains from utility privatization andregulation in Argentina, in: Public Policy for thePrivate Sector, December 1997, pp. 33-36.

Dnes, A.W. (1995), Franchising and privatization, in: PublicPolicy for the Private Sector, March 1995, pp. 5-8.

Estache, A. (1997), Designing regulatory institutions forinfrastructure—lessons from Argentina, in: PublicPolicy for the Private Sector, June 1997, pp. 13-16.

Foster, C.D. (1992), Privatisation, Regulation and theControl of Natural Monopoly, Oxford: Blackwell.

Guislain, P. (1997), The Privatization Challenge: A Strategic,Legal, and Institutional Analysis of InternationalExperience, The World Bank: Washington, D.C.

Izaguirre, A.K. (1998), Private participation in the electricitysector - recent trends, in: Public Policy for thePrivate Sector, December 1998, pp. 5-12.

Jones, L. (1994), Appropriate regulatory technology: Theinterplay of economic and institutional conditions,in: Proceedings of the World Bank AnnualConference on Development Economics, 1993,pp. 181-213.

Joskow, P. (1998), Regulatory Priorities for ReformingInfrastructure Sectors in Developing Countries,The World Bank Annual Bank Conference onDevelopment Economics, April 20-21, 1998,Washington, D.C.

Klein, M. and Irwin, T. (1996), Regulation water companies,in: Public Policy for Private Sector, March 1996,pp. 5-8.

Nells, J. and Roger, N. (1994), The Private SectorDevelopment Seminar: Increasing PrivateParticipation (http://www.seas.gwu.edu/student/damenam/telecom/conference/part8.htm).

Ogus, A. (1994), Regulation: Legal Form and EconomicTheory, Clarendon Press: Oxford.

Otobo, E. E. (1998), Privatization and regulation in Africa:Some key policy issues, in: DPMN Bulletin, Vol.V, No. 1, December 1998, pp. 23-28.

Paddon, M. (1998), Restructuring and privatization ofutilities in Asia Pacific region, in: Loretta de Luca(ed.), Labour and Social Dimensions ofPrivatization and Restructuring (Public Utilities:Water, Gas, Electricity), ILO: Geneva, pp. 43-108.

Plane, P. (1998), Privatization of water and electricity inAfrica, in: Loretta de Luca (ed.), Labour andSocial Dimensions of Privatization andRestructuring (Public Utilities: Water, Gas,Electricity), ILO: Geneva, pp. 1-42.

Schumpeter, J. (1987), Capitalism, Socialism and Democracy(6th edition), London: Unwin Paperbacks.

Sherer, F.M. (1980), Industrial Market Structure andEconomic Performance, Chicago: Rand McNally.

Silva, G. and Tynan, N. and Yilmaz, Y. (1998), Privateparticipation in the water and sewerage sector—recent trends, in: Public Policy for the PrivateSector, September 1998, pp. 41-48.

Slay, B. and Capelik, V. (1998), Natural monopoly regulationand competition policy in Russia, in: AntitrustBulletin, Spring 1998, pp. 1-17.

Stiglitz, J.E. (1998), Knowledge for Development. EconomicScience, Economic Policy and Economic Advice,Address to the World Bank's 10th Annual BankConference on Development Economics(ABCDE) (http://www.worldbank.org/html/extdr/extme/js-abcde98/js_abcde98.htm).

Teece, D. (1995), Telecommunications in Transition.Unbundling, Reintegration, and Competition, 1MICH.TEL.L.REV. 4, 1995 (http://www.umich.edu/~mttlr/VolOne/Teece.html).

UNCTAD (1995), Comparative Experiences withPrivatization: Policy insights and LessonsLearned, United Nations: New York and Geneva.

Vogel, S. (1997), International games with national rules:How regulation shapes competition in 'global'markets, in: Journal of Public Policy, Vol. 17, No.2, pp. 169-193.

Waddell, J. (1997), The Privatization of Monopolies, in:Economic Perspectives, Electronic Journals, Vol.2, No. 1, January 1997 (http://www.usia.gov/journals/ites/0197/ijee/ej5c4.htm).

Wanasinghe, S. and Wanasinghe, S. (1991), The return of theprivate sector to the road passenger transportbusiness: A case study of privatisation of bustransport in Sri Lanka, in: Geeta Gouri (ed.),Privatisation and Public Enterprise: The Asia-Pacific Experience, Oxford & IBH PublishingCo.: New Delhi, pp. 407-432.

World Bank (1997), The State in a Changing World, TheWorld Bank: Washington, D.C.

Yarrow, G. (1994), The Economics of Regulation, in: V.V.Ramanadham (ed.), Privatization and After:Monitoring and Regulation, Routledge: Londonand New York, pp. 35-46.

DESA Discussion Paper Series

No 1 Public versus Private Provision of PensionsBy Larry Willmore, December 1998

No 2 Inefficiencies of Global Capital MarketsBy Hugh Stretton, December 1998

No 3 Greening the National Accounts:Approach and Policy UseBy Peter Bartelmus, January 1999

No 4 Unpaid Work and Policy-MakingTowards a Broader Perspective of Work and EmploymentBy Joke Swiebel, February 1999

No 5 Trends in Consumption and Production:Selected MineralsBy Oleg Dzioubinski and Ralph Chipman, March 1999

No 6 Trends in Consumption and Production:Household Energy ConsumptionBy Oleg Dzioubinski and Ralph Chipman, April 1999

No 7 Promoting Sustainable Production and Consumption:Five Policy StudiesBy Tarcisio Alvarez-Rivero, Ralph Chipman and Erik Bryld, April 1999

DESA Discussion Papers are posted on the DESA web site:http://www.un.org/esa/papers.htm

Documents

Regulation policies concerning natural monopolies in developing