Levy Spiller _1994_ - The Institutional Foundations of Regulatory Commitment

The Institutional Foundations of Regulatory Commitment: A Comparative Analysis of

Telecommunications Regulation

Author(s): Brian Levy and Pablo T. Spiller

Source: Journal of Law, Economics, & Organization, Vol. 10, No. 2 (Oct., 1994), pp. 201-246

Published by: Oxford University Press

Stable URL: http://www.jstor.org/stable/764966

Accessed: 18/01/2009 18:36



Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at

http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you

may use content in the JSTOR archive only for your personal, non-commercial use.



Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at

http://www.jstor.org/action/showPublisher?publisherCode=oup.



Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed

page of such transmission.



JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the

scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that

promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.









Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of Law,

Economics, & Organization.









http://www.jstor.org

JLEO.V10N2 201







The InstitutionalFoundationsof Regulatory

Commitment:A ComparativeAnalysis

of TelecommunicationsRegulation

Brian Levy

The World Bank



Pablo T. Spiller

University of California, Berkeley



In this article we use transaction cost economics to analyze the determinants

of performance of privatized utilities in different political and social circum-

stances. Case studies are drawn from telecommunications regulation in Ar-

gentina, Chile, Jamaica, the Philippines and the United Kingdom. We explore

how political institutions interact with regulatory processes and economic

conditions in determining the potential for administrative expropriation or ma-

nipulation, and hence affect the sector's economic performance. We find that

performance can be satisfactory with a wide range of regulatory procedures,

as long as arbitrary administrative action can be restrained. We find also that

regulatory credibility can be developed in unpropitious environments, that

without such commitment long-term investment will not take place, that

achieving such commitment may require inflexible regulatory regimes, that in

some cases public ownership of utilities is the default mode of organization,

and furthermore, that it may be the only feasible alternative.



1. Introduction

This article provides a comparativeassessmentof the impactof core political

and social institutionson telecommunications and

regulatorystructures utility

performanceoutcomes in five countries:Argentina,Chile, Jamaica,the Phil-

ippines, and the United Kingdom.' Although in recent years institutional



BrianLevyis SeniorEconomist The World

at T.

Bank.Pablo Spilleris Visiting Professorof

BusinessandPublicPolicy,HaasSchoolof Business,University California,

of Berkeley. This

articleis partof the World Bankproject and

Regulation, Economic

"Institutions, Efficiency,"

with

undertaken partial support fromthe World BankResearch We

Committee. wouldlike to

thankLee Alston, Hadi Esfahani, AhmedGalal, Alice Hill, MattMcCubbins, RogerNoll,

Douglass North,MaryShirley, IngoVogelsang, BarryWeingast,Bjorn Wellenius,OliverWil-

and

liamson, twoanonymous for

referees helpful and and

comments suggestions, Janet Mowery

for editorial work.The opinionsexpressed hereareentirely and

thoseof the authors, do not

represent viewsof theWorld

the Bank,its Executive or

Directors, thecountries represent.

they

1. The country studieson whichwe baseouranalysis partof the World

are Bankproject

"Institutions, and

Regulation Economic Efficiency,"funded partby a grantfromthe World

in

BankResearch Committee. Because thenature thegrant, didnotinclude U.S. as one

of of we the

we

of ourcases.Although make references theU.S. inouranalysis,

to our

applying framework to

of

the evolution telecommunications and in

regulation performance the U.S. is left for future

research.

? 1994by Oxford Press.All rightsreserved.

University 8756-6222194/$5.00

202 The Journal Lav. Economics.& Organization, N2

of V10





analyses have made majorgains in accountingfor the conduct of regulatory

and other economic policies in the United States and elsewhere, the empirical

assessment of some key hypotheseshas been constrainedby the fact that over

long periods of time the U.S.'s core political institutions(institutionsidenti-

fied as key explanatorsof policy outcomes)have changedonly slowly.2 Cross-

country comparativeanalyses are one way to release this constraint.

Utilities privatizationand regulatoryreformare increasinglybilled as ways

to improve service quality and to lower prices. Here we argue that such

expectationsmay not always be attainable.In particular, looking at the prob-

lem of utilities regulation throughthe lens of transactioncost economics-

with its microanalytical perspective, its emphasison discriminating alignment

and remediableness, and its view of regulationas a contractingproblem-

provides an understanding the determinants performanceof privatized

of of

utilities in different political and social circumstances. Our objective is to

highlight how political institutions interact with regulatory processes and

economic conditions in exacerbatingor amelioratingthe potentialfor admin-

istrativeexpropriation manipulation,and hence determiningthe economic

or

performance of the sector.

We arguethatthe credibilityandeffectiveness of a regulatoryframework-

and hence its ability to facilitate privateinvestment-varies with a country's

political and social institutions. Further,we argue that performancecan be

satisfactory with a wide range of regulatoryprocedures, as long as three

complementary mechanisms restrainingarbitraryadministrativeaction are

all in place: (a) substantive restraints on the discretion of the regulator,

(b) formal or informal constraintson changing the regulatorysystem, and

(c) institutions that enforce the above formal-substantive or procedural-

constraints.Ourevidence suggests thatregulatorycommitmentcan indeed be

developed in what appearto be problematicenvironments,that without such

commitment long-term investment will not take place, that achieving such

commitmentmay requireinflexibleregulatoryregimes thatgo againstprevail-

ing academic views, that in some cases public ownership of utilities is the

default mode of organization,and furthermore, that such ownershipmay be

the only feasible alternative.

Political and social institutionsnot only affect the ability to restrainadmin-

istrative action, but they also have an independentimpact on the type of

regulation that can be implemented, and hence on the appropriate balance

between commitmentand flexibility. For example, relatively efficient regula-

tory rules (e.g., price caps, incentive schemes, use of competition) usually

require granting substantial discretion to the regulators. Thus, unless the

country's institutions allow for the separationof arbitrariness from useful

regulatorydiscretion, systems that grant too much administrative discretion







2. Several studies, however, have exploited the electoral changes and theirrepercussionson

the composition of the U.S. Congress and the executive in undertakingthose tests. See, for

example, Weingastand Moran (1983) and Spiller and Gely (1992).

TheInstitutional of

FoundationsRegulatory

Commitment

203





may not generate the high levels of investment and welfare expected from

private-sectorparticipation.Conversely, some countries have regulatoryre-

gimes that drasticallylimit the scope of regulatoryflexibility.Although such

regulatoryregimes may look inefficient, they may in fact fit the institutional

endowments of the countriesin question, and may provide substantialincen-

tives for investment.

Our analysis may be especially relevantfor the design of regulatorypolicy

in developing, newly industrializing, and previously socialist countries,

where lack of economic developmentmay be relatedto a generalizedlack of

administrativerestraints.3But the results are also relevant in understanding

the historical evolution of utilities regulation and ownership in developed

countries. Indeed, as the analysis of the U.K. case highlights, restraining

regulatorydiscretionseems to be behind the developmentof regulatoryinsti-

tutions in developed countries as well.4

A word of caution is in order, though. Our effort at comparativeanalysis

has some importantlimitations. The need for detailed analyses of the politi-

cal, social, and regulatoryinstitutions of each country naturallylimits the

number of cases we are able to analyze.5 Consequently, we do not offer

formal statisticaltests of our centralpropositions(such tests would requirea

methodology not feasible at this stage of our research)but insteadprovide an

analytical frameworkand casual but systematicallycollected and researched

evidence. While we believe that our results provide strong supportto some

core propositionsaligned with the new institutionalism,the spiritof the paper

is exploratory-as much an exercise in hypothesisformulationas in hypothe-

sis testing.



2. The Analytical Framework

2.1 The Problemof Utilities

Three special featurescharacterizeutilities and provide the startingpoint for

our analysis. First, most (but not all) utility services are characterizedby

importanteconomies of scale and scope. Second, most utilities' assets are

highly specific and non-redeployable(althoughthe extent of sunk investments

varies with the applicationand with technology). Third, utility services typ-

ically have a broad range of domestic users, usually overlappingthe voting

population of the country. Viewed throughthe lens of the new institutional

economics, these characteristics

createcontractingproblemsthatundercutthe

ability of ordinarymarketmechanismsto deliver first-bestperformance.6



3. For an extremely interesting analysis of the role of central government discretion in

determiningincentives for private-sectorinvestment, see Weingast(1993).

4. While the McNollgast approachdoes not say it in those words, this is indeed their main

message concerning, for example, administrativeprocedures in the United States. See, for

example, McCubbins, Noll, and Weingast(1987).

5. Multiple, distinct regulatoryepisodes within individual countries do, however, provide

additional"degrees of freedom."

6. See, among others, North (1990), Williamson (1988), Goldberg (1976), and Barzel

(1989).

204 The Journalof Law,Economics,& Organization, N2

V10





Economies of scale and scope and highly specific assets imply that the

numberof providersof basic utility services is going to be relatively small.

Because a large proportionof the utilities' assets are sunk, a utility will be

willing to operateeven if it cannot recover its sunk investmentsas long as it

covers its operating costs.7 Widespreaddomestic consumptionimplies that

the pricing of utilities is always going to be political.8Furthermore, whiff

the

of monopoly (particularlywhen ownershipis concentratedin foreign hands)

increases the gains from political action.

The combination of significant investments in durable, specific assets

with the high level of politicization of utilities has the following result: util-

ities are highly vulnerableto administrative of

expropriation their vast quasi-

rents. Administrativeexpropriationmay take several forms. Although the

easiest form of administrativeexpropriationis the setting of prices below

long-run average costs, it may also take the form of specific requirements

concerning investment, equipment purchases, or labor contract conditions

that extract the company's quasi-rents. Where the threat of administrative

expropriation great, privateinvestorswill limit theirexposure.9Thus, coun-

is

tries where administrative discretionis the norm may find that public owner-

ship of utilities arises because the hazardsof direct privateinvestmentare so

great.10

Politics and technology stronglyinteractin shapingthe potentialfor admin-

istrativeexpropriation. Sectors with very small sunk investmentsor with rapid

asset depreciationwill not be prime candidatesfor administrative expropria-

tion.1 Similarly, administrativeexpropriationmay not be expected in the

short run in countries or jurisdictions experiencingrapid growth in the de-

mand for utilities services, as the costs of investmentdelays may limit the

political gains from opportunisticbehavior. In the longer run, however, the

fight for control over the institutionsof government, and the corresponding





7. Observe that financing requirementsfor sunk investments are not part of the operating

costs. Inability to repay debt requiredto pay sunk investments will only bring the utility to

bankruptcy, it still will not be worth liquidating.Asset liquidationwill take place only when

but

operating revenues fall short of operatingcosts, where operatingcosts include a returnon the

nonspecific investmentsof the enterprise.

8. This does not mean thatthe pricingof othersectors, not characterized large economies

by

of scale and sunk investments, is not going to be political as well (e.g., bread prices in poor

countries). What we mean is that the politicization of utilities' pricing, together with its asset

characteristics,bring about unique problems.

9. Investors may limit their exposure by selecting technologies that-although they imply

lower qualityand higheroperatingcosts-may requirelower levels of specific investments(e.g.,

cellular ratherthan fixed-link telephony).

10. Observe, though, that public enterprisesare also subject to the forms of administrative

expropriationwe are consideringhere. In this case, though, systematicunderfunding will be the

norm.

11. For example, setting the prices of bananasbelow their long-runaveragecost will have at

most a one-yeareffect on bananaprices. On the otherhand, settingprices of waterservices below

long-run average costs may imply a reductionin prices over quite a long period of time.

The Institutional Foundations of Regulatory Commitment 205







exposethe utilitiesto potential

divisionof the spoils, will continuously ad-

or

ministrative outright expropriation,even in rapidly

growing economies.



2.2 Resolutionof the RegulatoryProblem:A Framework Empirical

for Analysis

for

This section lays out a generalframework empirical,country-specific

analysisof the extentto whichthe regulatory outlined

problems abovehave

beenresolved,explains why theresolution theformit did,anddiscusses

took

the relationbetweenregulatory outcomesand the performance private

of

utilities.Subsequentsectionsfill in the details.



2.2.1 RegulatoryDesign: Governance and Incentives. In trying to under-

standdifferent to

abilities commit particular

countries' to regulatory

processes

andinstitutions, findit usefulto look at regulation a designproblem.

we as

Regulatorydesign has two components:regulatorygovernance and regula-

We

toryincentives. definethe governance structure a regulatory

of systemas

that

the mechanisms societiesuse to constrain regulatory discretion to

and

to

resolveconflictsthatarisein relation these constraints.'2 regulatory

The

incentivestructure the

comprises rulesgoverningutilitypricing,cross- or

directsubsidies,entry,interconnection, In contrast regulatory

etc. to gover-

nance,the structure regulatory

of incentives beenthe central

has preoccupa-

tionof virtually theoretical

all workon regulation. mainresult thisstudy,

A of

though,is thatsuchemphasis inadequate.

is we

Although findthatregulatory

incentivesindeed affect performance, their impact(positiveor negative)

comesto theforefront if regulatory

only has

governance successfully beenput

intoplace.13

Bothregulatory and are

governance incentives choicevariables thehandsin

of policy-makers. choices,however, constrained

The are ones. Choicesas to

regulatory are

governance constrained thespecificinstitutional

by endowment

of the nation,whichbothdetermines formandtheseverity the regula-

the of

and

toryproblems shapesthe rangeof optionsavailable resolvingfor them.

Choicesas to regulatory are

incentives alsoconstrained a nation's

by specific

institutionalendowment. a

Moreover, nation's choiceof the governance fea-

turesof the regulatory systemwill havean independent effecton the typeof

regulatory incentives areviable.

that



2.2.2 Institutional

Endowment Governance. Following North

and Regulatory

(1990) and others, we define the institutional of

endowment a nationas

five

comprising elements. First,therearea country's and

legislative executive





12. Williamson would call such constraintson regulatorydecision-making"contractual gover-

nance institutions"(see Williamson, 1985:35).

13. Commenting on the interactionamong technology (institutions),governance, and price

(regulatorydetail), Williamson (1985:36) says: "[i]nasmuchas price and governanceare linked,

partiesto a contractshould not expect to have theircake (low price) and eat it too (no safeguard)."

In other words, there is no "free institutionallunch."

206 TheJouralof Law, & V10

Economics, Organization, N2





Thesearetheformal

institutions. mechanisms forappointing

(a) and

legislators

and

decision-makers for makinglaws and regulations, apartfromjudicial

decision-making; for implementing

(b) theselaws;and(c) thatdetermine the

relations and

betweenthe legislature the executive.Secondarethe country's

Thesecomprise formal

judicialinstitutions. its mechanisms forappointing

(a)

the

judges and determining internal of

structure the judiciaryand (b) for

impartiallyresolving disputesamong private parties, between

or private parties

andthestate.Third customs other

are and informal broadly

but accepted norms

thataregenerally to the

understood constrain actionof individuals institu-

or

tions.Fourth thecharacter thecontending

is of socialinterests

withina society

andthebalance between the

them,including roleof ideology. Finally, thereare

theadministrative of

capabilities thenation. of

Each theseelements subject

is to

change,andthe determinants eachcanalsobe thesubject study.In this

of of

analysis, though, we treateach of these elementsas exogenous-as the institu-

tional endowmentof a particular nation. Here we highlightthe first two, and

comment on the remainingthree whenever appropriate.

The form of a country'slegislative and executive institutionsinfluencesthe

nature of its regulatory problems. The crucial issue is to what extent the

structureand organizationof these institutionsimpose constraintsupon gov-

ernmentalaction. The range of formal institutionalmechanismsfor restrain-

ing governmentalauthorityincludes the following: explicit separationof pow-

ers between legislative, executive, and judicial organs of government;14 a

written constitution that limits the legislative power of the executive and is

enforced by the courts; two legislative houses elected under differentvoting

rules;15an electoral system calibrated to produce either a proliferationof

minorityparties or a set of partieswhose ability to impose discipline on their

legislators is weak;16and a federal structureof power, with strong decentral-

ization even to the local level.17 Utility regulationis likely to be far more

credible-and the regulatoryproblemless severe-in countrieswith political





14. For analysis of the role of separationof powers in diminishing the discretion of the

executive, see Gely and Spiller (1990) and McCubbins, Noll, and Weingast(1987, 1989) and

references therein.

15. Nonsimultaneouselections for the differentbranchesof governmenttend to create natural

political divisions and thus electoral checks and balances (see Jacobson, 1990). For an in-depth

analysis of the determinantsof the relative powers of the executive, see Shugart and Carey

(1992).

16. Electoral rules also have importanteffects on the "effective numberof parties"that will

tend to result from elections and, thus, on the extent of governmentalcontrolover the legislative

process. For example, it is widely perceived that proportional representation tends to generatea

large numberof parties, while first-past-the-post with relatively small districtelections tends to

create bipolar party configurations. This result has been called Duverger's Law in political

science (see Duverger, 1954). More generally,see Taagepera Shugart(1993). For analysesof

and

how the structure political partiesdependson the natureof electoralrules (with applicationsto

of

the U.K.), see Cain, Ferejohn, and Fiorina (1987) and Cox (1987).

17. On the role of federalism in reducing the potential for administrativediscretion, see

Weingast(1993) and references therein.

The Institutional of

Foundations Regulatory

Commitment 207





systems that constrain executive and legislative discretion. Note, however,

thatcredibilityis often achievedat the expense of flexibility.The same mecha-

nisms that make it difficult to impose arbitrary changes in the rules may also

make it difficultto enact sensible rules in the firstplace, or to efficientlyadapt

the rules in the face of changingcircumstances.Thus, in countrieswith these

types of political institutions, the introductionof reforms may have to await

the occurrenceof a drastic shock to the political system.

Legislative and executive institutionsmay also limit a country'sregulatory

governanceoptions. In some parliamentary systems, for example, the execu-

tive has substantialcontrol over both the legislative agenda and legislative

outcomes.18 In such countries, if legislative and executive powers alternate

between political partieswith substantiallydifferentinterests,specific legisla-

tion need not constitute a viable safeguardagainst administrative discretion,

as changes in the law could follow directly from a change in government.19

Similarly,if the executive has stronglegislative powers, administrative proce-

dures and administrativelaw by themselves will not be able to constrainthe

executive, who will tend to predominateover the judiciary in the interpreta-

tion of laws. In this case, administrative proceduresrequiresome base other

than administrativelaw.

A strong and independentjudiciary could serve as the basis for limiting

administrative discretionin several ways. For example, the priordevelopment

of a body of administrative law opens the governanceoption of constraining

discretionthroughadministrative procedures.20 Also, a traditionof efficiently

upholding contractsand propertyrights opens the governanceoption of con-





18. WVhile parliamentary systemsgrantsuch powersin principle,whether they do so in

practicedepends uponthenature electoral

of rulesandthepolitical party system.Parliamentary

systems whose electoralrules bring aboutfragmented legislatures would not providethe

executive-usuallyheaded a minority

by partywitha coalition builton a very narrow of

set

specific common interests-withmuch scopeforlegislative initiative. contrast,

By electoral

rules

thatcreatestrong two-party parliamentarysystems-as well as someotherkindsof nonparlia-

mentary politicalinstitutions-would grantthe executivelargelegislative powers.Foran in-

depth discussion thedifference

of between and

parliamentary presidential and

systems, theroleof

electoral rulesin determining relative

the of

power theexecutive, Shugart Carey

see and (1992).

19. In the U.K., regulatory have

frameworks traditionally evolved a

through seriesof actsof

For

Parliament. example, major regulation

gas was in

legislation passed 1847,1859,1870,1871,

1873,and 1875.Similarly, waterregulation was

legislation passedin 1847, 1863, 1870, 1873,

1875,and 1887.Systematic of

regulation electricity companies in

started 1882,onlyfouryears

aftertheinauguration thefirst

of publicdemonstration lighting a public

of by The

authority. 1882

act was followedby majorlegislation 1888, 1899, 1919, and 1922, culminating the

in with

Electricity Act

(Supply) of 1926,whichcreated Centralthe Electricity For

Board. discussions of

theevolution utility

of in

regulation theU.K., see Dimock (1933),Hormell (1928),Keen(1925),

Self andWatson (1952),andSpillerandVogelsang (1993b).

20. Thistraditionally beenthewayadministrative

has discretion restrained theU.S., as

is in

regulatory statutes havetended be quitevague.Foran analysis thechoiceof specificity

to of of

see

statutes, Schwartz, Spiller,andUrbiztondo (1993).Observe, that

however, administrative

law may not developin a systemwherethe executivehas strongcontrolover the legislative

process.

208 The Journal Law.Economics,&Organization. N2

of V10





strainingdiscretion throughthe use of formalregulatorycontracts(licenses).

This option is particularlyvaluable for countries where the executive has a

strong hold over the legislative process. Further,a traditionof judicial inde-

pendence and efficiency opens the governanceoption of using administrative

tribunalsto resolve conflicts between the governmentand the utilitywithin the

contours of the existing regulatorysystem. Finally, it provides assurances

against governmental deviation from specific legislative or constitutional

commitmentsthat underpinthe regulatorysystem. While there are no simple

ways of measuringjudicial strength,two featuresseem to us to be key deter-

minants: the extent of perceived judicial corruptionand whether the courts

have a history of deciding against the government. A corruptjudiciary that

attemptsto be independentmay easily find its corruption publicly disclosed,

with a clampdown following.21 Similarly, a judiciary that seldom decides

againstthe government,even on contractdisputes, cannotbe interpreted be to

strong.22



2.2.3 Institutional Endowmentsand RegulatoryIncentives. If the regulatory

incentive structure is to promote welfare, it should facilitate investment,

allocatively efficient pricing, and the introductionof new services and tech-

nologies. Yet regulatoryincentives cannotbe implementedin an institutional

of

vacuum. The country'sinstitutionalendowment, the character distributive

politics, and the nature of its regulatorygovernance structureall affect the

potential for the successful design of regulatoryincentives.

Consider first the constraintsthat the nation's institutionalendowmentim-

poses on the design of regulatoryincentives. Administrative capabilities-the

ability of the nation'sprofessionals(e.g., academics, lawyers, bureaucrats) to

handle complex regulatoryconcepts and processes in a relatively efficacious

manner,without triggeringexcessive disputes and litigation-are of particu-

lar relevance. These capabilitieswill determinethe potentialfor the successful

implementation complex regulatorydesigns. Thus, regulatorysystems that

of

call for complex implementationwill be inadequatein nations with weak

administrativecapabilities.

As for the impact of distributivepolitics, it can constrain the extent of

allocative efficiency that can be achieved by regulatoryincentives. A major





21. While a corruptjudiciary may tend to side against the governmenton contractdisputes

(after all, governmentlawyers will have very little incentive to bribejudges, while such incen-

judiciarybecomes an easy hostage

tives clearly exist for privatelawyers), on key issues a corrupt

to the government. Unexpectedpress disclosuresmay triggerwidespreadscandals, loss of legit-

imacy, and even calls for judicial reformand sanctions. Considerthe following conclusion of a

WorldBank reporton Bolivia: "In additionto being slow, courts are also perceivedto be corrupt.

In a 1991 survey of 226 litigants, 48 percent claimed to have made illegal payments to court

personnel. Confidencein the courts is also weakenedby thejudiciary'sperceivedlack of indepen-

dence. Judges' independenceis compromisedby a historyof executive supremacy,the country's

traditionof political patronage,and the proceduresfor the selection, promotionand discipline of

judges" (WorldBank, 1992:43).

22. This is a necessary but not a sufficientcondition for strength.

The Institutional of Commitment 209

Foundations Regulatory





feature of telecommunicationsdistributivepolitics is the demand for cross-

subsidization from businesses to residential users, which has translatedin

many countries into very high prices for long-distanceor international calls.

Unless eliminated, such cross-subsidizationlimits the potentialfor exploiting

the advantagesof competition.

Finally, and perhaps more important, the institutionalrealities of some

countries may be such that resolutionof the governanceproblemsconstrains

the range of regulatory incentive options to third (or even fourth) best. In

some countries, the only way to constrainadministrative arbitrarinessmay be

to almost totally withdraw administrativediscretion. The resulting gover-

nance considerationswould constrainthe rangeof workableregulatoryincen-

tive designs to those that provide only limited flexibility. In telecommunica-

tions, where technological change is rapid, the cost of this reduction in

regulatoryflexibility is high-but so too is the cost of failing to adequately

restraindiscretion.

Yet, even given these constraints, the individual country studies suggest

that in most cases thereexists a broadrangeof discretionfor policy-makersas

to the design of regulatorygovernanceand incentivestructures.Indeed, utility

performanceappears to be best in countries that have achieved a good fit

between their exogenous institutions and their regulatory governance and

incentive designs, and worst in those instances where regulatorydesign pro-

ceeded without attentionto the exogenous institutional realities. Furthermore,

the qualityof design of regulatoryincentivesemerges in the countrystudies as

an independentinfluence on performance.23



2.3 A Decision Tree

The decision tree in Figure 1 summarizesthe above discussion and sets the

stage for analysis of the five countriesin a way thathighlightsthe impactof a

country's exogenous institutionalendowmenton its regulatorydesign.24The

heavier lines in the figure are branchesrepresentedby observations in our

sample. Because our frameworkhas implicationsfor other institutionalenvi-

ronments, we present the decision tree in its general formulation.We make

four sets of distinctions.

The first distinction is between those countriesthat have domestic institu-

tions capable of credibly refrainingfrom arbitrary administrative action and

those that do not. Our frameworksuggests that the existence of an indepen-

dent judiciary with a reputationfor impartiality,and whose decisions are

enforced, is a necessaryconditionfor makingthese crediblecommitments.As

discussed furtherbelow, among the countriesstudied, Chile, Jamaica,and the



23. Indeed, it is possible (although we have no examples where this is the case) that an

especially poorly designed regulatoryincentive structurecould in itself underminethe credibility

of a regulatorysystem, even if the regulatorygoverance foundationsappearfirm.

24. The discussion that follows can be interpreted eitherpositive or normative.It is positive

as

in that it predictswhich combinationsof backgroundinstitutionsand regulatorysystems will lead

to good performance which will not. It is normativein that it suggests whatkind of regulatory

and

design will be credible given the backgroundinstitutionsof the country in question.

210 The Journal Law,Economics,& Organization, N2

of V10





noGlFt

ADECSKONTRE









LUNFE



YES





JUDIARY?









OovsNERSr

$ERWAONAL

GUARANTES









1. institutional

Figure A decisiontreemodelof a country's endowments.





United Kingdom had in place a well-functioningjudiciary, while Argentina

and the Philippines did not. Countrieslacking a well-functioningjudiciary

will face difficulties in the short term in developing a regulatory system

capable of sustainingefficient levels of privateparticipationand investment,

scarceresourcesto such

and thereis little reasonfor them to devote substantial

an effort. Instead, alternativemechanismsof securingcommitment(like inter-

national guarantees)will be necessary.

Among countries capable of using regulation to secure commitment, the

second distinctionis between those countrieswhose electoral, legislative, and

executive institutions are structuredin a way that enables them to achieve

credible regulatorycommitmentsvia legislation and those countriesthat can

best achieve crediblecommitmentsby embeddingtheirregulatorysystems in

the operatinglicenses of privatecompanies. As we discuss below, legislation

may provide regulatorycredibility in political systems that do not generate

unified governments, like presidential systems with multichamberlegisla-

tures, fragmentedparties, and nonsimultaneouselections. For legislation to

provideregulatorycredibility,however, it should make specific the process by

which regulatorydecisions are to be taken. Otherwise, regulatorydiscretion

will be unchecked. Legislation, however, does not provide credibility to

regulatory policy in, for example, two-party parliamentary systems where

parties alternatein government. In this case, regulatorycredibility can be

obtainedby basing the regulatoryprocess in contractlaw, ratherthanadminis-

The Institutional Foundations of Regulatory Commitment 211







trativelaw.25As will be seen, Chile falls into the formercategory(as does the

United States), and Jamaicaand the United Kingdom into the latter.

The third distinction, not depicted in Figure 1, is between countries that

requirespecific, substantiverules to achieve credibilityand countriesthat can

use flexible regulatoryprocesses and still restrainarbitrary action. In general,

the potential for flexibility in design will be highest in two sets of countries.

The first set comprises those whose exogenous institutionalendowments in-

law

clude informalnormsor bodies of administrative thatrestrainthe arbitrary

use of governmentalpower even if explicit legal restraintsare absent. The

second set comprises those where an institutionalized process of argumenta-

tion and consensus formation sets de facto limits on the extent to which a

private utility can be subjected to administrativeexpropriation.Among the

countriesstudied, althoughneitherJamaicanor the U.K. can providecredibil-

ity throughlegislation, Jamaicafalls into the category of countriesrequiring

specific, substantive rules to achieve regulatorycredibility, while the U.K.

can achieve such credibility with a more flexible regulatoryprocess. Chile,

while able to use legislation as its regulatoryinstrument,also has the ability to

achieve credibility with a more flexible regulatoryframework.

Among countries that require very specific, substantiverules to achieve

credibility,the fourthand final distinctionin Figure 1 is between those coun-

tries that have strong administrative capabilitiesand those that do not. Coun-

tries with strong administrativecapabilities can put in place a regulatory

system based on specific, substantiverules thatcan both attractinvestmentby

restrainingarbitraryaction and promote efficiency and flexibility. Countries

with weak capabilities may have to settle for less efficient rules for their

regulatorysystem to work. Among the countriesstudied, Chile falls into the

former category, and Jamaicathe latter.

Viewed from a differentperspective, the decision tree identifiesthree com-

plementarysets of mechanismsto restrainarbitrary action. First are substan-

tive restraintson regulatory discretion, which can take the form of either

process regulation or specific, substantiverules (i.e., distinctions 3 and 4).

Second are restraintson changing the regulatorysystem, of eitherthe legisla-

tive or licensing variety (i.e., distinction 2). Finally, there are institutions,

notably a judiciary,for enforcing both the substantiverestraints restraints

and

on system changes (i.e., distinction 1). A centralhypothesisto be exploredin

the comparativecountryanalysis is thatfor privateperformance be satisfac-

to

tory all three mechanismsmust be in place and they must be properlyaligned

with the specific characterof a country's backgroundinstitutions.





25. When electoral laws are designed to returnthe same party to power in every election,

complex partydecision-makingmay provideregulatorycredibility,even in the absence of de jure

contractual arrangementsor very precise legislation. This seems to have been the Japanese

example, until the 1993 elections. See Baron (1992) for an excellent discussion of how decision-

making in Japan'sLiberal Democratic Party (LDP) was structuredso as to provide for a large

numberof veto points.

212 The Journalof Law.Economics.& Organization, N2

V10





3. The Institutional Background of the Country Cases

This section describes the exogenous variables-the institutionalbackground

of each of the five countries.26Section 4 then providesthe link between these

and our endogenousvariables-regulatoryregimesand regulatory performance.

Table 1 provides a snapshot of the main institutionalcharacteristicsof

Argentina, Chile, Jamaica, the Philippines, and the United Kingdom, and

relatesthem to the potentialfor opportunistic governmentbehavior.In explor-

atory researchof the kind describedhere, there is an ever-present risk of ex

post rationalization interpreting role of exogenous variables.To obviate

in the

this risk, we have tried as much as possible to anchorour depictions of the

exogenous institutionsof the five countries in prior scholarly writings. Fur-

ther, we alert the readerwhereverwe push our interpretation the behavior

of

of a country's institutions beyond previous interpretations.We divide the

countries into three groups: traditionalparliamentarysystems (Jamaica and

the U.K.), the archetypalpresidentialsystems (Chile), and rent-seeking presi-

dential systems (Argentinaand the Philippines). Each group has distinct im-

plications for what type of regulatoryregimes are workable.



3.1 Traditional Parliamentary Systems: Jamaica and the United Kingdom

Jamaicaand the U.K. provide the observationsfor the Figure 1 branchthat

answers yes to the "unifiedgovernment"question. The political systems of

Jamaicaand the U.K. are parliamentary, characterized a strongjudiciary

by

and electoral rules that tend to generate two strong parties.27As a conse-

quence, the majorityparty invariablyhas an absolute majorityin Parliament

and controls both the governmentand the legislature.28The two countries'

judicial institutionsare also similar.29Britainhas led the world in the develop-

ment of a judiciary with an exceptional reputationfor probity. Jamaica's

judiciary is also well regarded, and has ruled against the government on



26. The detailed case studies on which this article is based follow a research framework

developed in Spiller and Levy (1991). That frameworkemphasizes, as this article does, the

contractingproblemsendemic to utilities regulation,and requiresthe gatheringof detailedinfor-

mation about the regulatorystructure(includinginformation regulatoryproceedingsand court

on

cases) over time. It also requires the collection of a common set of informationconcerning

investments, prices, and outputs.

27. JamaicainheritedBritish political institutionsupon obtainingits independencefrom that

country in 1962. In Jamaica, the percentageof the vote and the numberof seats obtained by

candidatesindependentof the two main parties(the PNP and the JLP) fell, respectively,from 35

percentand 5 (out of 35 seats) in 1944 to I percentand 0 since the 1959 election (see Stone, 1981,

1986). See also Cox (1987) and Cain, Ferejohn, and Fiorina (1987) for an analysis of party

politics in the U.K.

28. It is not surprisingthat both countrieshave had drasticpolicy shifts as partiesalternatein

power.

29. While Jamaica'sjudicial institutionsparallel those of the U.K., there are two important

differences. First, as a former colony, Jamaica'sjudicial system has retainedthe right of final

appeal of its decisions to the Privy Council in the U.K., providingan unusuallystrong source of

credibility.Second, unlike the U.K., Jamaicahas combinedthe institutionof a sovereign parlia-

ment with a written constitution.

The Institutional Foundations of Regulatory Commitment 213







numerousoccasions.30 Note, though, the implications for the judiciary of a

political system based upon a sovereign two-partyparliamentwith no written

is

constitution:undersuch a system, the parliament free to act as it pleases, so

long as it follows generallyacceptedprocedures(on which more later),but the

courts have no formalrestrainingauthority.31 a large extent because of this

To

feature of parliamentary politics, in neither country has the judiciary devel-

law

oped a strong administrative doctrine, althoughboth have a long contract-

law tradition, and have upheld contracts against the government.32In both

countries, then, administrativedecisions are essentially undertakenin the

dark.33Agencies do not necessarilyjustify theirdecisions (althoughthe U.K.

Office of Telecommunications Regulationdoes),34nor necessarilyhold public

hearings.35'36 Judicial review of agency decisions is not customaryin either

country.37 in

Furthermore, the U.K., regulatoryreview is not a strongweapon

for the regulatedfirms and intervenors,38 because regulatorscan make deci-

sions that effectively preventjudicial review.39

Thus, the formal institutionsof governmentin the U.K. and Jamaicaallow

for substantial governmental discretion. This governmentaldiscretion can





30. For example, duringthe firstadministration PrimeMinisterMichael Manley,the courts

of

blocked Manley's attemptto expropriatelandholdingsby requiringthe governmentto pay fair

compensation.

31. The fact that the U.K. does not have a writtenconstitution,though, has not precludedthe

development of a large body of constitutionallaw.

32. See Salzberg (1991) for a discussion of the secondary role of the U.K. judiciary in

statutory interpretation.See Spiller and Sampson (1993) for a discussion of the judiciary in

Jamaica,and Spiller and Vogelsang(1993b) for a discussion of the role of thejudiciary in license

enforcementuntil the nationalizationof the electricity sector after WorldWarII.

33. For a discussion of the development of regulatoryagencies and their proceduresin the

U.K. since the beginning of the privatizationprocess, see Veljanovski(1991).

34. For, example, "the Cable Authorityis not bound to give reasons for its decisions and it

may well decide not to spell out its selection criteriain any detail in an effort to retainthe greatest

flexibility" (Baldwin and McCrudden, 1987:292).

35. The JamaicanPublic Utilities Commission (JPUC) during the period 1966-75, did hold

public hearings and was, in principle, subject to U.S.-style conditions for judicial review.

36. See Baldwin and McCrudden(1987) for a discussion of procedures in various U.K.

regulatory agencies, including the Office of Fair Trading and the Monopolies and Mergers

Commission.

37. Judicial review in the U.K., however, is becoming more frequent,in particular the local

at

governmentlevel, as conflict between local and centralgovernmenthas increased. Baldwin and

McCrudden(1987:57), quoting Sir Michael Kerry,the former TreasurySolicitor, observe that

judicial review of administrative agencies has "increasedfrom a handfula year in the 60s to 50-

100 in the early 70s, to a rateof about400 a year in the firstsix monthsof [1982]." Furthermore,

Kerry reported that most of these cases "come under two main heads, applications to quash

planning decisions . . . and immigrationcases." (see Kerry, 1983:168; and Young, 1985).

38. However, in the U.K. judicial review may be effective in resolving disputes among

branchesof government. See Baldwin and McCrudden(1987:59) for a discussion of this issue.

39. In the U.K. judicial review can be initiated on proceduralgrounds or if the regulator's

decision has been unreasonable.A claim of unreasonableness, however, rarelywins, because the

regulatoris not requiredto provide a thoroughexplanationof the basis for making a particular

decision (e.g., the granting of a cable license). See Baldwin and McCrudden(1987:292-93).

Chile, Jamaica, Philippinesand United Kingdom

Featuresof Argentina,

Table 1. Institutional



Legislativeand

Country ExecutiveInstitutions Judicial Institutions Social Conflict



Argentina Separationof powers Weak and politically Basic urban-rural R

with fragmented manipulable. conflict.

partystructure.

Chile Separationof powers Strongand Regional balance. R

with fragmented professional;

partystructure. defended private

Orderlyaltemation propertyduring

of powerexcept socialist period.

duringGen.

Pinochet'sregime.

Jamaica Parliamentarysystem Strong and pro- Small middle class R

withtwo strong fessional; upheld holds political

parties. utilitycontracts balance. No major

against government regional differences.

priorto 1966 and Violenturbanpoor.

defended private

propertyduringthe

first Manley

administration.

Philippines Separationof powers Weakand corrupt. Regional competition R

with weak parties among members of

and strong elite.

presidency. Until

1960s, partial

United States

supervision. Military

regime during

Marcos.

United Kingdom Sovereign parliament Verystrong and Large middle class, S

with strong two-party professional; holding political

system. upholds contracts balance.

against the

government.Weak

administrativelaw.

216 The Journal Law.Economics,&Organization, N2

of V10





translateinto importantpolicy shifts, with the judiciaryplaying a minor role

in restrainingadministrative discretion. Informalrules of legislative decision-

making, however, restrain both governments' ability to undertakepolicy

shifts without consulting interest groups. For example, over the years the

U.K. has developed decision-makingprocesses that tend to limit the endemic

secrecy of parliamentary system governmentsby allowing interest-group con-

sultations.Althoughboth these countrieshave the potentialto undertake large

policy shifts, some amount of consensus in policy-makingis required.The

need for consensus-makingproceduresarises as much from the need to legiti-

mize governmentaldecision-makingas from the need to avoid cracks in the

governing party with the consequent loss of legislature support. Ways to

develop consensus and conduct interest-group consultationsvary from coun-

try to country;they include the role and use of ad hoc commissions and white

and green papers,40 use of collective ministerial

the as

responsibility,41 well as

the use of party organizationsto settle disagreementsprior to parliamentary

action. Within-partyconsensus-buildingmeasuresare useful for single-party

governments not just because they grant stability and credibility to their

actions, but also because they provide legitimacy to the party leadership.

Legitimacy is importantin safeguardingagainstother factions' attacksand in

maintainingbackbenchers'support.42





40. For example, majorpolicy changes startwith an ad hoc commission that is chargedwith

analyzing the policy change the governmentis interestedin undertaking. The outcome of the ad

hoc commission is a consultative"greenpaper,"which while not representing government's

the

exact intentionsis neverthelesscirculatedfor public comment. (Telecommunications reform, for

example, started in 1981 with the Beesley report, which recommendedunrestrictedresale of

leased lines, British Telecommunication'sprice flexibility on leased lines, and networkentry.)

Following this comment period, the government may decide to drop or to continue (with or

without modification)the policy initiative. If the governmentdecides to go aheadwith the policy

change, it will eventually publish a "white paper," which forms the basis for the cabinet's

legislative initiative. This process, which is also followed in Jamaica,has the virtueof providing

interest groups with warning signs about impendingpolicy changes. Interestgroups then can

lobby partymembersand legislators in an attemptto influencethe policy change in theirdesired

direction. See Miller (1985:197-211) for a discussion of how the decision-makingprocess in the

U.K. providesopportunities interestedpartiesto participate the legislativeprocess. See also

for in

Spiller and Vogelsang(1993a). The role of this decision-makingprocess is not too dissimilarfrom

the advance warningfeaturesof the U.S. Administrative ProcedureAct. For a discussion of the

U.S. AdministrativeProcedureAct, see McCubbins,Noll, and WVeingast (1987) and Spiller and

Urbiztondo(1993).

41. On the role of ministerialresponsibilityin limiting the potential for breaks in the party's

controlover government,see Cox (1993). Cox's mainpoint is thatthe effect of requiring joint and

several ministerialresponsibility"is to increasethe powerof the executive body by decreasingthe

chance that executive decisions will be challenged in any broaderarena. Indeed, the convention

of collective responsibilityin the United Kingdom has been so effective in preventinginternal

Cabinet disagreements from spilling out into the Commons that many have questioned the

independentimportanceof the legislature"Cox (1993:47).

42. To some extent, because Japan'sLDP until recently had continuouscontrol over govern-

ment in a contestable legislative environment, it developed a formal process of government

decision-makingprior to submittinglegislation to the Diet. This process involves the formal use

of the LDP's Policy Affairs ResearchCommittee(PARC) and industryadvisorycommittees, as

The Institutional of Commitment 217

Foundations Regulatory





In the U.K., furthermore, the level of policy implementation,the West-

at

minster model of governmentdefines the prerogativesand responsibilitiesof

politiciansand bureaucrats,43 mutuallyrecognizedboundaries authori-

with of

ty, even where theirpower is not explicitly constrained.Jamaicaand the U.K.

differ, however, in the extent of professionalismof the bureaucracy. While the

U.K.'s bureaucracy professionaland largely apolitical, the politicizationof

is

of

the bureaucracyduring the first administration Jamaica'sPrime Minister

Manley seems to have had an irreversibleeffect in reducingthe qualifications

and role of its bureaucracy.44

Thus both the U.K. and Jamaicahave electoral systems that provide for

great legislative flexibility,andjudiciariesthatdo not stronglyrestrainadmin-

istrativeactions. As a consequence, neithercan base its governancestructure

on legislative acts. They need furtherconstraints.As we will see, both found

them in contractlaw.



3.2 The ArchetypalPresidentialSystem: Chile

Chile, with its archetypalpresidentialsystem,45is the countryin our sample

that provides the observationsfor the Figure I branchthat answers no to the

"unifiedgovernment"question. Between 1973 and 1989, Chile went through

a period of militaryrule underGeneralAugusto Pinochet. This period, how-

ever, stands out as a 15-year exception to more than 100 years of rule by

civilian government. For the rest of this 100-year period, the country was

governed by a constitution that embodied the principles of separation of

powers, orderly transferof authority,and regularelections between compet-

ing parties. For an extended period, voters were divided among multiple

parties, with none strongenough to legislate, except by coalition. Legislators'

led

independencefrom centralpartyapparatus to the developmentof a legisla-

A

ture with a strong sense of local representation.46 series of constitutional



well as informal consultations with Zoku (tribe) Diet members and LDP factions. For an ex-

tremely insightfuldescriptionof governmentaldecision-makingin Japan,see Baron(1992:chap.

13). Lupia and McCubbins (1993) stress that a major role of PARC's review is to provide

backbencherswith informationabout Cabinet intentions. Because the Japaneseelectoral system

develops a strongerpersonal vote in Japanthan in the U.K., Japanesefactions are more intrin-

sically independentfrom the party leadershipthan in the U.K. As a consequence, more formal

consensus-building processes are required to keep interfactionconflict from spilling into the

legislature.

43. For an interesting transaction-costdiscussion of the raise of the Westminstermodel of

bureaucracy,see Palmer (1992).

44. During his first administration,Prime MinisterManley broke with political traditionand

replacedmany PermanentSecretaries.Since then, Permanent Secretarieshave lost some of their

influence and prestige, reducing the position's attractivenessto highly qualified middle-class

individuals (see Stone, 1986). For a discussion of the U.K.'s bureaucracy, and a comparisonto

that in the U.S., see, for example, Fesler (1983).

45. Shugartand Carey (1992:173-74) use Chile's political system until 1970 to characterize

the "archetypalpresidentialsystem."

46. At the same time, concurrentelections providedfor the developmentof bipolarcoalitions

thatoffered voterschoices over nationalpolicies withoutthe need to develop strongcentralparties

(Shugartand Carey, 1992:174).

of V10

218 The Journal Law.Economics,&Organization. N2





in

reformsbetween 1958 and 1973 shiftedthe balanceof authority favorof the

presidency and elevated national (ratherthan regional) politics and parties to

center stage. These reformsgrantedmore executive powers at the same time

thatthey developed strongerchecks andbalances.47 The reformsincreasedthe

potential for conflict between the legislative and the executive branches,

which became evident during the regime of President Salvador Allende.

Thus, while not entirely resistantto extreme pressure, Chile's long-standing

set of legislative and executive institutionsand the natureof its checks and

balances can be seen as potentially providing some credible safeguards

against arbitrarychanges in the regulatoryregime governing utilities. This

period of extended stability, broken only by the military takeover in 1973,

developed in Chile a strong respect for institutions.48

Chile's strong, professional, and independent judiciary provides a partic-

ularly effective check on the government,on issues of both constitutionaland

statutoryinterpretation. example, Allende's 1970-73 governmentrepeat-

For

edly clashed with the courts over issues of expropriation and compensation,

with the courts refusing to back down. Its large pool of highly qualified

professionals,has served Chile well in providinggovernmentsfrom the differ-

ent parties access to qualifiedpolitical appointees.

Chile's diffuse political power, then, provides opportunitiesfor designing

regulatorygovernancestructures along several alternativelines. Because spe-

cific legislation is more difficult to change in Chile than, say, in the U.K.

specific legislation may play a more importantrole in the regulatorygover-

nance structureof Chile than in the U.K. On the other hand, Chile's strong

judiciaryprovides also for implementingregulatorygovernancebased purely

on proceduresor contractlaw.49



3.3 Rent-Seeking PresidentialSystems: Argentinaand the Philippines

Argentinaand the Philippinesprovidethe observationsfor the Figure 1 branch

that answers no to the "independent judiciary"question. These two countries

share some common political features. First, both have modeled their formal

political institutions upon those of the United States, creating a complex



47. Shugartand Carey (1992:183) call the system before the constitutionalchanges of 1958

and 1970 the "inefficientsecret," because the regionalizationof politics and the relatively weak

powers of the president provided for a very dynamic legislature. Concerning the post-1970

Chilean system, Shugartand Carey (1992:200) describe it as the strongest among presidential

systems. It featuresa strong germanenesslimitationon amendmentsand permitsonly the presi-

dent to offer a budget bill. Thus, the presidentfully controlsthe budgetprocess, althoughthat is

not the case on nonbudgetaryissues.

48. One may ask whetherthis respectfor institutionspersistedeven duringthe militaryregime

of General Pinochet. Even then signs of respect for long-standinginstitutionswere evident. For

example, during the military regime the constitutionwas amended by a popular referendum.

Similarly,the militaryregime's decision to returnthe countryto a democraticsystem followed an

electoral loss.

49. See, McCubbins, Noll, and Weingast(1987) for a discussion of the role of proceduresin

providing commitment.

The Institutional of

Foundations Regulatory

Commitment 219





system of checks and balances.50Indeed, both countries have experienced

quite extendedperiods of constitutionalrule duringwhich executive authority

the

was largely restrained.Yet structureis not determinative: history of both

countrieshas been punctuatedby periods of extra-constitutional in which

rule

governmentalauthority was concentratedin the president(see Hill and Ab-

dala, 1993; and Esfahani 1993).51Second, in both countries,the endemic lack

of respect for constitutionalorder has translatedinto a corruptbureaucracy

and judiciary, and into turs to the militaryas the panaceafor interest-group

conflicts.52,53



both have

50. Forexample, countries separation powers of betweenlegislature,executive,and

judiciary; written

a constitution and

limitinglegislative executive powerandenforced the by

two

courts; legislative houses,withweakpolitical and

parties; a federal structure power,with

of

strongdecentralization to thelocallevel.Although

even rulesaresuchthattheeffective

electoral

number parties

of contesting in

elections thePhilippines 2, in Argentina is higher 2 and

is it than

increasing, 3.4 and

reaching in 1989(Shugart Carey, To

1992:220-21). a large extent,thiscanbe

by of

explained theimportance nonconcurrent elections Argentina. factthattheeffective

in The

is

number parties thePhilippines twodoesnotmean thesearestrong

of in that parties.Indeed,in

as

thePhilippines, in theU.S., parties notcontrol

do nor

nominations, do theyuse list ranks to

of

control order elections poolthevotes.Thus,as in theU.S., Filipino

the or are

legislators quite

of

independent theircentral parties.

51. Argentina's 1853 constitutional separation of powers remained in place for almost a

century, until it was amended in 1949 by President Juan Per6n in a way that substantially

expanded the powers of the president. The military overthrewPer6n in 1956 and restored the

earlier constitution. However, two subsequentmilitarygovernmentscame to power in 1966 and

in 1976. In the past, during Argentina'speriods of civilian rule, governmentseither have been

fragmented,making it difficult to enact sustainedeconomic reforms, or have pursuedzero-sum

policies which rewardedthe supporters those in power. The presentadministration President

of of

Menem is an important exception to this patternof civilian rule. Not only did his PeronistParty

win control of both houses of Congress(as well as the executive) in the 1990 elections, but when

Menem took power, Congress (with the active supportof Menem'sopponents)ceded him special

powers to deal with the economic crisis (see Hill and Abdala, 1993).

52. The theory to explain why some polities develop rules of the game that defer the arbitra-

tion of interest-group conflicts to the judiciaryand the legislatureratherthanto the armyis yet to

be developed. Calvert's (1992) view of constitutionsas conventions is a step forward, in that

coordinationgames have multipleequilibria,with one equilibriumbeing characterized lack ofby

coordination. We do not attemptto provide an answer to this question here.

53. In Argentina, the disdain for constitutionallegality appearseven in the transferof power

following democratic elections. The transferof power from the Radical Party to the current

PeronistPartygovernmentwas advancedto several monthsbefore the legal date. The transferof

power was arrangedthroughan agreementbetween the PeronistPartyand the Radical Partythat

required the incumbent Congress (not yet dominated by the Peronists) to pass an Economic

EmergencyLaw grantingthe futurepresident(CarlosMenem) the power to implementby decree

a series of economic measures (see Hill and Abdala, 1993). Also, membersof the judiciary are

appointedby the country'spresident(and confirmedby the Senate)and serve for life, giving them

substantialformal independence.In practice,however, changes in government(whetherconstitu-

tional or nonconstitutional) have in the past been accompaniedby turnover amongmembersof the

judiciary. Indeed, in 1990 the new government of President Menem, with the agreement of

Congress, altered the SupremeCourtby increasingthe numberof justices from five to nine and

naming all new justices. By 1993 PresidentMenem was able to appointsix of the nine Supreme

Courtjustices. The Menem Court has recently been plagued by corruptionscandals (see The

Miami Herald, October 12, 1993).

220 The Journal of Lawv,

Economics. & Organization. V10 N2







Although the governments of both the Philippines and Argentina have

traditionallyfollowed "beggar-thy-neighbor" policies toward political oppo-

nents, the governing Filipino elites have followed, even during the Marcos

regime, a "nonexpropriation" norm, wherebyassets of political opponentsare

not expropriatedoutright (Esfahani, 1993). Instead, since control over the

governmentrotatesamong elite groups, business owners not aligned with the

elite in power tend to be subjectto administrative (see

expropriation Esfahani,

1993).

To summarize, both Argentina's and the Philippines' institutionalframe-

works seem to be sufficiently weak, and the scope for executive dominance

sufficiently strong, that our frameworksuggests that regulatorygovernance

has to go beyond the choices of legislation, procedures,or contractlaw.54 We

discuss these issues in Section 4.2.



4. Exogenous Endowments and Regulatory Commitment Mechanisms

This section describes and interpretsthe regulatoryexperience of the five

countries in question. The discussion of regulationis organizedaroundthe

three mechanismsto restrainarbitrary administrativeaction that were identi-

fied earlier:(a) substantiverestraints the discretionof the regulator

on thatare

written into the design of a regulatorysystem, (b) restraintson changing the

regulatory system, and (c) institutions for enforcing both the substantive

restraintsand restraintson system changes. Each subsectionsets the stage for

discussing regulationby briefly describing the evolution of telecommunica-

tions ownership in the relevantcountries.

As the analysis reveals, across the five countries and different periods,

regulatory systems worked and attractedprivate investment at reasonable

rates of returnonly when all three mechanismswere in place. Moreover,the

evidence suggests that the exogenous institutionalendowmentsof individual



In the Philippines, on the other hand, while its writtenconstitutionincorporatesan independent

judiciary and prior to the proclamationof martial law in 1972 the judiciary was reasonably

independent, the Philippines press has long been filled with tales of judicial scandal. Under

martial law, its independence was furtherconstrained and President Marcos effectively was

empoweredto remove any judge. Since 1986, the judiciaryhas regainedsome independenceand

influence, althoughcorruption scandalscontinue. For example, in 1992 the Philippines'Supreme

Courtblocked the entry of a competing international communicationscarrier(EasternTelecom).

A few months later the author of the decision (Justice Hugo Gutierrez) resigned, following

allegations that the opinion had been written by a PLDT lawyer (Panaligan, 1993-cited in

Esfahani, 1993).

54. Viewed over the longer term, though, the multiplicityof partiesin both countriessuggests

that if the currentdemocraticepisodes representa clear break with unconstitutional transfersof

power, future authority may be weaker than at present. U.S.-style legislative and executive

institutions were reasonably effective in constrainingarbitrarygovernmentalauthority in the

Philippines in the first quarter-century after it obtained independencefrom the United States in

1946-although the Philippines'constitutiontilted the balanceof power more sharplytowardthe

executive than does the United States model (see Abueva, 1988). PresidentMarcos'sdeclaration

of martiallaw in 1972 (despite some modest reformsafter 1978), furtherconcentratedpower in

the executive. Since his overthrow in 1986 the country has revertedto a somewhat modified

version of its pre-martial-law political institutions.

of

Foundations Regulatory

The Institutional Commitment 221





countries have influencedboth whetherthey were able to put in place work-

able governance designs for their regulatorysystem and, if so, the specific

forms of such designs. Tables 2 through4 summarizethe informationpre-

sented below.



4.1 Traditional Parliamentary Systems: Jamaica and the United Kingdom

Table2 summarizesthe chronologyof telecommunications ownership,regula-

tion, and performancein Jamaicaand the U.K.55 Three featuresof the chro-

nology are especially noteworthy and are examined in some depth below.

First, both the U.K. and Jamaicahave had sustainedperiodsof high levels of

private investment in telecommunications:in Jamaicabefore 1962 and after

1987, in the U.K. since corporatizationand privatization(1982 and 1984,

respectively). Second, duringthe periods of high investment,regulationtook

differentforms in the two countries. In Jamaica,rate-of-return regulationhas

been the standardregulatory method, although with importantdifferences

between the pre-1962 and post-1987 periods;in the U.K., on the other hand,

price-cap regulation was introduced for the first time in the regulation of

BritishTelecommunication (BT). Finally,althoughbothJamaicaand the U.K.

had periods in which private ownership was associated with strong invest-

ment, Jamaica experienced a long decade (from the early 1960s to the

mid-1970s) during which private investmentlagged.

In this section we attemptto explain these three featuresin the light of the

frameworkdeveloped earlier.The thrustof this section is the following: First,

despite seeming differences in their regulatoryframeworks,the strong com-

monalitiesin the two countries'regulatorygovernancestructures have roots in

the similarities of their exogenous institutionalendowments;second, the dif-

ferences in the substantiverestraints developed in both countriescan be traced

to importantdifferences in their respective exogenous endowments;third, in

all three high-investmentperiods the regulatorysystems were appropriately

aligned with exogenous endowments in a way that createdworkablemecha-

nisms to restrainarbitrary action; finally,Jamaica'sdisastrous 1966-75 expe-

rience was the consequence of a misalignmentbetween its exogenous institu-

tions and the chosen regulatorygovernance structure.



4.1.1 The High-Investment Periods. As Table 2 describes, Jamaicahas had

two high-investmentperiods:the firstcontinueduntil the government'sdeci-

sion (in 1962) to switch from license-based, rate-of-return regulation to a

U.S.-style, Public Utility Commission-based regulatory framework. The

second period startedwith the creation and privatizationof Telecommunica-

tions of Jamaica (TOJ) in 1987. The U.K.'s high-investmentperiod began

with BT's corporatizationand subsequentprivatizationin 1984. As we dis-

cuss below, these events provide the observationsfor the Figure I branchthat

answers yes to the "de jure contractualarrangement" question.

Although the regulatoryincentive schemes in the three periods are quite



55. This section is based on Spiller and Sampson (1993) and Spiller and Vogelsang (1993a).

Table2. Regulatory

History Mechanisms:Jamaica and the United Kingdom

and Commitment







Private Substantiv

Country/Period Ownership RegulatoryHistory Performance Restraints



Jamaica-1966 Privatelyowned Rate of return(7- Sustained network Monopoly righ

with separate 9%)on assets expansion up to plus precise

domestic and stipulatedin the 1962. Steady, rate of retur

international license. Ad hoc moderate rate of long period;

companies.a rate boards re- return. writtenin the

sponsible for license.

rate reviews.

Decisions chal-

lengeable in

courts.

Jamaica In 1962 govern- New license is- Initialnetworkex- rate

"Fair" of

1966-75 ment announced sued in 1966 pansion, then return.

intentionto specifies JTCto slowdown.Low

change regula- be regulatedby profitability.

tion system at li- JPUC. License

cense renewal specifies rate of

time. Sale of return be fair

to

JTCto CTC and reasonable,

(Continental although Tele-

TelephoneCom- phone Act spe-

pany) in 1966. cifies 8% return.

In 1971 govern- JPUC interprets

menttakes ma- legislative man-

joritystake in date to be not

new internation- binding. JPUC

al operations institutesrate in-

company,and in creases based

1975 takes over on investment

JTC. performance.

Sustained regu-

latoryconflict.

Jamaica Creationof Tele- License stipulates Majorinvestment Monopoly righ

1987- communications rate of return in domestic net- plus precise

of Jamaica (17.5-20%) on work. Reduction rate of return

(TOJ),a joint shareholders're- in real prices long period;

venturebetween valued assets. and sustained writteninto

the government Speedy arbitra- high profitability. license.

of Jamaica and tion. Govern-

Cable and Wire- ment cannot

less, to combine challenge in-

JTCand JAM- vestments. Gov-

INTEL. Privati- ernmentto

zation starts in monitorquality,

1988. and may require

quality improve-

ments.

United King- Publiclyowned License specifies of

Takeoff invest- Main regulator

dom 1984- from 1912 until price cap, and a ment in 1982, sues, includ

privatization in complex mecha- with large gains RPI-X, with a

1984. nism to resolve in national specific X, w

regulatorycon- welfare. ten into licen

flicts









company:Cable and Wireless

alnternational (WestIndies);domestic company:JamaicaTelephone

Company(JTC); shareho

majority

Table3. RegulatoryHistoryand Commitment

Mechanisms:Chile







Substantive

Period Ownership RegulatoryHistory Performance Restraints



1930-58 Controlledby ITT Regulatorylaw Initially,relative Vague. Based on

(International open-ended, no rapid expansion, legislation.

Telephoneand independent reg- then slowdown.

Telegraph). ulatory.Govern-

ment had rightto

intervenein the

company'sopera-

tions. Disputes to

be resolved by

SupremeCourt.

1958-73 Partialtakeoverof Regulatorysub- Networkexpansion Strengthenedby

CTC (Compaiia agreements in accelerates prior 1958 subagree-

de TelIfonosde 1958 and 1967 to intervention ments.

Chile) by govern- eliminated and nationaliza-

ment in 1967, vagueness con- tion.

governmentinter- cerning rate of

ventionin 1971, relturn(set now

and total take- at 10% instead of

over in 1974. EN- up to 10%)and

TELcreated to imposed new in-

provide long- vestment obliga-

distance services tions.

in 1964.

1987- Privatizationin Detailed bench- Unprecedentedly Veryexplicit,pre-

1988. mark regulation high rates of net- cise price regul

for noncompeti- workexpansion tion based on

tive marketseg- and trafficgrowth rate of returnof

ments. Price subsequent to efficientfirm,

regulationbased privatization. plus explicit pro

on performance cess for definin

of theoretically noncompetitive

efficient firm,re- activities subjec

calibrated every to price regula-

5 years with inter- tion. Both stipu-

im indexation. lated by law.

Law stipulates

process to be fol-

lowed to compute

and rate of return

to efficient firm.

Explicit arbitra-

tion procedures

for recalibration.

Regulatorydeci-

sions can be ap-

pealed to

Supreme Court.

Table4. Regulatory

Historyand Commitment

Mechanisms:Argentinaand the Philippines







Substantiv

Country Ownership RegulatoryHistory Performance Restraint



Argentina Maintelephone License-based regula- Highly profitable Explicit price-

companies na- tory framework spe- to initialinves- adjustmentf

tionalizedin cified initial2 years tors. Investment mula.

1946, to forma based on rate of re- levels match or

nationalcompa- turn,followedby 5 exceed license

ny (ENTEL). Di- years withprice cap renewal require-

vided into two withx = 0. Exclusive ments. Very

regionalcompa- license to be ex- high prices (in

nies and pri- tended to 10 years US$), and quali-

vatizedin 1990. only if minimum in- ty problems re-

vestmenttargets are main. Too soon

met. Infact, price to determine full

freeze instituted 4 impact on wel-

monthsafterprivat- fare.

ization. Companies

negotiated indexa-

tion based on U.S.

consumer price in-

dex. Regulatory

agency created un-

der the Ministry of

Economy.Appeals

only to the Minister.

Regulatoryframe-

workcurrently under

review.

Philippines Privatesince in- Long-standingregula- Alternationbe- None.

ception, with tion by commission tween stagna-

ownershipshift- with vague mandate, tion and periods

ing fromGTEto lack of politicalin- of moderate in-

private Filipino dependence, and vestment, fol-

parties in 1967. modest power. lowing political

cycle. Very high

unmet demand.

un-

Profitability

known.

of Economics. Organization, N2

228 TheJournal Law, & V10





dissimilar (Jamaica having in both periods a rate-of-return regulation with

specific profitabilitytargets and strong limitations on regulatorydiscretion,

while the U.K. has no formalprofitability targetsin its price-capregime), the

regulatory governance structuresare quite similar. The licenses served as

contractsbetween the regulatedcompaniesand their respectivegovernments,

since attemptsby the governmentsto deviate from the license specifications

could be, and were, challenged in court. For the judiciary to serve as a

credible arbitrator, however, the licenses must be specific enough thatjudicial

outcomes are predictable.Licenses, then, can serve as an effective restraint to

administrative action. Indeed, in the three episodes, licenses have been very

specific about price-setting procedures. Both Jamaicanlicenses require that

prices provide the companywith a specific rate of return(7-9 percenton real

operating assets in the pre-1966 period, and 17.5-20 percent on revalued

net

shareholders' worthsince 1987). The 1987 license also limits the abilityof

the regulatoryto disqualify investments or to delay price increases through

administrativesilence. Furthermore,the 1987 license provides for speedy

arbitrationto resolve price-settingconflicts. Both Jamaicanlicenses specifi-

cally grantedthe company the ability to challenge governmentdecisions in

court. BT's 1984 license (and all its amendments)is equally specific aboutits

price-settingprocedures.Price-settingpowers are grantedto BT, with only ex

post supervisionby the regulatorto see that regulatedprices follow the price-

cap structure specified in the license. In both countries, attempts by the

regulatorto alter the main featuresof the regulatoryregime requiresa license

modification. In Jamaica,license modificationsrequirethe agreementof the

company.In the U.K., license modificationsinvolve a well-specifiedprocess,

which to be undertakenagainst BT's will, requires the agreement of the

Monopolies and MergersCommission(MMC) and the Secretaryof Tradeand

Industry(the Board of Trade). Furthermore,an attemptby the regulatorto

circumventsuch a process can be challenged in court.

Thus, in both Jamaica and the U.K. the use of very specific licenses

provided the companies with substantialassurancesabout the expected prof-

itability of their investments.56Furthermore,because these licenses were

specified for long periods of time, were not easily modified, and were en-

forceable in the courts, the companies were assuredof substantialregulatory

stability.57







56. Although BT's license was very specific on price setting, it was quite vague on quality

regulation. Thus, Oftel, the telecommunicationsregulator,has spent most of its efforts, and

administrativediscretion, in implementingquality-control measures(see Spiller and Vogelsang,

1993a). Given BT's disastrous internalquality-controlsystems at privatizationtime, it is not

altogether clear that such efforts by Oftel were not in the interests of BT's shareholdersand

customers.

57. Jamaica'slicenses traditionallyhave been grantedfor 25 years. While BT's license was

grantedfor a similar period, its currentprice-capfeatureexpires after 5 years. Thus, unless BT's

license is modified before the sunset provision of its price-capregulation,BT's pricingbecomes

The Institutional of

Foundations Regulatory

Commitment 229





law

The emphasis on contractratherthan administrative in providingregu-

latorycredibility in both Jamaicaand the U.K. is consistentwith the natureof

their political institutions.Their systems of parliamentary sovereignty imply

thatthe mere existence of a law governingregulation,even if it is very precise

and specific, providesvery little assurancethatthe currentregulationswill not

be subsequentlymodified. On the other hand, the courts in both Jamaicaand

the U.K. have a strong history of upholdingcontractsamong privateparties.

Jamaicancourts were called to resolve contract(regulatory)disputesbetween

the government and the regulated company before 1962, while the U.K.

courts dealt with electric power license interpretation issues prior to World

War II. While U.K. courts have yet to hear and decide a case against a

regulatorof the privatizedutilities, it seems that the threatof legal action has

No

restrainedthe UnitedKingdom'sregulators.58.59 disputesover the termsof

the 1988 Jamaicanlicenses have yet been broughtto court.

To summarize, in the threeperiods with substantialprivateinvestment,the

three regulatoryregimes specified importantsubstantiverestrictionson the

discretion of the regulator;furthermore,there were serious impedimentsto

changing the regulatoryregimes;and finally,in all threeepisodes therewas an

institutionthat could credibly be called upon to enforce both the substantive

restrictionson the regulatorand the stability of the regulatoryregime.



4.1.1 Differences in RegulatoryGovernance. Although both the U.K. and

Jamaica have based their respective regulatorygovernance structureson li-

censes, BT's license has a measureof flexibility that TOJ's license does not

have. BT's license can be amendedagainst its will; TOJ's cannot. Amending

BT's license against its will, however, is no minortask. It requiresagreement

among the Director General of Telecommunications,the Monopolies and

Mergers Commission, and the Secretaryof Tradeand Industry.These three

authorities can be expected to differ with one another in their regulatory

views; indeed, two of them (the Director General and the Monopolies and

MergersCommission) enjoy more than a modicumof independencefrom the

governmentof the day. As Spiller and Vogelsang(1993a) show, the traditional





unregulated.This sunset provision gives BT a strong bargainingposition, since undertakinga

license modificationagainst its will may take as much as a full year (see Spiller and Vogelsang,

1993a).

58. Spiller and Vogelsang (1993a) report that following the failure at the Monopolies and

Mergers Commission of the DirectorGeneral'sproposal for license reformconcerningthe mar-

keting of 900-type calls, the Director General was advised not to proceed with his proposal

because he would be successfully challenged in courts.

59. Recently, however, Mercuryfiled a suit against BT and Oftel and it won its preliminary

court hearing. Mercuryclaimed that "Oftelhad not offer it with reasonableterms for the carriage

of its trafficby BT, througha consistentmisinterpretation BT's governmentlicence. Mercury's

of

victory . . . means that there will be a full hearingof its applicationfor a legal interpretation

of

that part of the licence which deal with interconnection" (Financial Times, 1 March 1994, our

italics).

230 The Joural of Law.Economics,& Organization, N2

V10







independence of the MMC and the inability of the Secretaryof Trade and

Industryto dictate decisions to the DirectorGeneralof Telecommunications60

imply that the multiplicityof veto points make a license amendmentagainst

BT difficult, though not impossible.

Jamaica's licenses have never had that measure of flexibility. Since the

beginningof utilityregulation,Jamaica'slicenses have required agreement

the

of the company for any amendment.61 That requirement given the regu-

has

lated company substantial bargainingpower. As a consequence, even dramat-

ic technological or political changes may fail to bring about changes in the

regulatory environment.62While such rigidity could be criticized from a

purely normativeperspective,Jamaica'sinstitutional endowmentprovidesthe

clue for the need to introduceinstitutionalrigidities in the regulatorygover-

nance. In the U.K., regulatoryflexibility was built on a foundationof a long

traditionof informalchecks and balances and the use of independent,expert

commissions. By contrast, Jamaicahas never had a system of administrative

checks and balances. The JamaicanPublic Utilities Commission created in

1966 was the first, and only, experimentwith independentcommissions, and

as the next section shows, the result was quite disastrous. Furthermore, the

repercussionsof the politicizationof the bureaucracy duringthe first adminis-

trationof Prime Minister Manley in the early 1970s eliminatedthe potential

for using delegation of regulatoryresponsibilitiesas a source of commitment

(Spiller and Sampson, 1993).63Our frameworksuggests, then, that attempt-



60. Constitutionalinterpretation traditionallyhas seen independentregulatoryagencies in the

U.K. as separate entities from their respective Secretary of State. The (CAA) revocation of

Laker's Skytrain license to fly to the U.S. in 1976 following the issuance of the Secretaryof

State's guidance for the CAA was successfully challenged in the Court of Appeal (see Laker

Airvays v Departmentof Trade [1977] QB 643, cited in Baldwin and McCrudden, 1987:167).

The Courtstatedthat the Department Trade'sguidance"couldsupplementthe CAA's statutory

of

objectives; it could not replace them. .... In issuing peremptoryinstructionsto the CAA it

constituted direction ratherthan guidance.... Parliamentcould not have set up an elaborate

licensing code, subject to limited powers of direction, only to allow the Crownto renderlicences

useless by use of the prerogativepower." See Baldwin and McCrudden(1987:167). This, how-

ever, did not preclude Ministerialinfluence. Indeed, because the statuteprovidedfor appeal of

CAA's decisions to the Secretaryof State, since Laker 18 appealsto the Secretaryof State have

been successful, in whole or in part, while priorto Lakerno appealto the Secretaryof State was

successful (Baldwin and McCrudden,1987:168). WhatLakershows is thatthe Secretaryof State

may use the powers given by statute and consistent with the constitutionalview of the role of

independentagencies.

61. As mentioned,Jamaica'slicenses traditionally have been grantedfor 25 years. At expira-

tion, the drawingof a new license agreementwould also requirethe acquiescenceof the company.

62. For example, the U.S. experience with flexible competitiveboundariesas a way to adapt

the regulatoryregime to new technologies (see Knieps and Spiller, 1983) cannotbe implemented

in Jamaica without compensating the company for the reduction in its profitability.Thus, for

example, if new entrantswere to be allowed, interconnection rateswould have to compensatethe

company for its profit loss. This is not altogether different from the position taken by New

Zealand's High Court in its recent Clear Communications decision concerningthe interconnec-

tion of new entrantsto New Zealand Telecom's network (see Mueller, 1993).

63. As mentioned above, the politicization of the bureaucracyduring the mid-1970s dimin-

ished the attractiveness the position in the eyes of the Jamaicanmiddleclass. Severalcommen-

of

The Institutional Foundations of Regulatory Commitment 231









ing to design in Jamaicaa regulatorygovernance system with as much flex-

ibility as that in the U.K. will most probably undermine the credibility

of Jamaica's regulatory process, and hence negatively impact the sector's

performance.



4.1.2 Differencesin RegulatoryIncentives. Jamaicaand the U.K. differ not

only in their regulatorygovernance structuresbut also in the natureof their

regulatory incentive structures. Jamaica pursued a straightforward rate-of-

returnregulatorysystem,64 while the U.K. introducedprice-capregulation.

Because the normative inefficiencies of rate-of-returnregulation had been

discussed at length in the economic literatureby the early 1980s and because

price caps were already widely discussed as superiorto rate of returnby the

late 1980s,65 it is proper to question whether Jamaica'sregulatorysystem

introducedin the late 1980s was the result of a mistake, capture, or of the

constraintsimposed by the institutionalbackgroundof the country.

Although we cannot reconstructthe mental processes by which Jamaica's

government privatized TOJ and compare it to that undertakenby the U.K.

government, it is clear that in both instances substantialemphasis was given

to how to design the regulatory system to be applied to the privatized

company-in the case of the U.K., through the commissioning of several

white papers,66and in the case of Jamaica,throughnumerouscabinet meet-

ings.67 While captureand mistakes are difficult to discardas reasons for any

regulatory decision,68 our framework sees the institutionalbackgroundin

Jamaica as limiting the potential implementationof a flexible price-cap re-

gime.69 In the U.K., the regulator'srevision of the price-cap rules is con-





tators have mentioned to us that since the 1980s secretarieshave tended to depend less on the

advice of their permanentsecretariesand more on non-civil-service advisors.

64. The modem Jamaicanrate-of-return system, though, differsin important dimensionsfrom

that common in the U.S. or Chile prior to the 1970s, and even from Jamaica'sown pre-1966

system, in that the regulatoris given very little latitudeto challenge capital investmentsand the

rate of returnis based on stockholders'revaluedassets ratherthanon fixed or operatingassets (as

was the case pre-1966).

65. For analyses of price-cap regulation, see, for example, the RANDJournal of Economics

symposium on price caps (Vol. 20, No. 3, 1989).

66. For a discussion of the legislative process up to the passage of the TelecommunicationsAct

of 1984, see Newman (1986:8-I 1). See also Spiller and Vogelsang (1993a).

67. Conversationwith RichardDowner of Coopers& Lybrand,Jamaica,and consultantto the

Jamaicangovernment.

68. Spiller and Sampson (1993) provideevidence againstthe capturehypothesisby computing

the value of TOJ under its license scheme and comparing it to the price paid by the private

investors. They find that the price of TOJ was very close to theircomputationof TOJ value as of

privatizationdate. In theircomputationthey assumeda 4 percentdividendpolicy and thatthe rate

of returnon equity would be the lower bound specified in the license. As the discount rate, they

use the real returnof Jamaicanbonds.

69. However, it is possible to say thatthe regulatorysystem in Jamaicahas a de facto price-cap

feature, because a freeze has been imposed on the price of local calls since privatizationwhile

international calls have been used to provide the necessary rate of return.A five-year freeze was

informallyagreedupon at privatization time. Since then, TOJhas maintainedthe nominalfreeze.

V10

Economics, Organization, N2

232 TheJouralof Lavw &





strained by a very specific process that is built around the integrity and

credible independence from government of the various layers of authority

required to approve license amendments and around a range of informal

norms in the U.K., which ensures that the process cannot easily be manipu-

to

lated. Such a process, however, could not easily be transferred Jamaica.

Jamaicalacks the multiple layers of independentauthoritynecessary to build

in credible U.K.-style restraints.70 Since price caps are by naturetransitory,71

it is reasonableto expect the X factor to be increasedopportunistically the

at

next price-cap renewal. As a consequence, the regulatedcompany will limit

its exposure, unless the initialprice-capregime would provideit with substan-

tial up-frontrents.72In other words, U.K.-style RPI-X (price cap) regulation

is not readily transferable.



4.1.3 RegulatoryMisalignment: Jamaica 1966-1975. In 1962 the Jamaican

governmentdecided to move, at license renewaltime, from a license-basedto

a PUC-style regulatorystructure.In 1966 the governmentissued new licenses

to the domestic company based on the JamaicanPublic Utilities Act of 1966,

which mandated"fair"rates of return. It also established a permanentand

independentregulatorycommission, the JamaicanPublic Utilities Commis-

sion (JPUC), to oversee regulation of domestic telephone and electricity

services. The JPUC had substantialadministrativediscretion and promoted

participation the regulatoryprocess by a wide rangeof interestgroups. This

in

episode provides the observationsfor the Figure 1 branchthat answers no to

the "de jure contractualarrangement" question.

From its inception, the relationshipbetween the JPUC and the domestic

company (JamaicaTelephoneCompany,or JTC) were acrimonious,and the

conflict increased over time. Price increases fell substantiallybehind infla-

tion, and the JPUC introduceda policy of promising price increases condi-

tionedon qualityimprovements investment

and levels. The domesticoperations

company stopped all investmentprogramsin 1962 following the government

decision to change the regulatory regime. Except for a short period fol-

lowing the U.S.-based ContinentalTelephone Company's takeover of do-

mestic telephone operationsin 1966, there was no network expansion until

the 1980s.

The primaryreason for the acrimonyand the deteriorating performanceof

JTC was the absence of substantiverestraintsin the regulatorysystem. The

JPUC was created to replicate the operationsand freedoms of a U.S.-style

PUC. The JPUC invited participationfrom interest groups, undertookhear-





70. Indeed, since the eliminationof the JamaicanPublic Utility Commission(see below) there

has not been even a single regulatory agency with substantial visibility and reputationfor

independence.

71. A fixed price cap in perpetuity will bring widely unstable profitabilityoutcomes, and

hence is not credible.

72. We claim this to be the case in Argentina(see below).

The Institutional Foundations of Regulatory Commitment 233







ings, and made determinations withoutstrongconstraints.The enablinglegis-

lation only requiredit to set "fairand reasonable"rates. Although the JPUC

Act of 1966 grantedthe Jamaicancourtsthe right to review JPUC'sdecisions,

during the 1966-75 period the judiciary did not provide any importantre-

straintto the JPUC. Indeed, given the vagueness of the license, the "unrea-

sonableness"criterionrequiredto overturna governmentaldecision became a

very stringent standard to overcome.73 The JPUC was, for all purposes,

unchecked, except for the direct intervention of the government.74Thus,

duringthis period, the judiciary could not be called to provide the regulatory

credibility inherentin the pre-1966 and post-1987 regulatoryregimes.



4.2 The ArchetypalPiesidential System: Chile

Table 3 provides a summarychronology of telecommunicationsownership,

regulation, and performancein Chile.75 As in Jamaica, Chile's regulatory

history can be separatedinto high- and low-investmentperiods. During the

high-investmentperiods, regulationtook very differentforms-rate of return

in 1958-71, and benchmarkregulation from 1987 on. The low-investment

periods were (a) from the end of WorldWarII until the late 1905s and (b) from

the Allende government,throughthe interventionin 1971 and nationalization

in 1974, until the regulatoryreformsof the mid-1980s. Here we focus only on

the period under private ownership (up to 1971 and from 1987 on). In this

section we attemptto explain and interpretthese patternsin the light of the

frameworkdeveloped earlier. The thrustof this section is as follows: first,

during the periods of high investment, the regulatoryschemes were better

aligned with Chile's exogenous endowmentsthan duringthe low-investment

period; second, that such alignments restrainedarbitrary action in the three

dimensions noted above.



4.2.1 The Low-Investment Period. The Compafifa de Tel6fonos de Chile

(CTC) was operating in Chile as early as 1880. In 1930 it entered into a

contract with the government (later written as Law No. 4791/30), which







73. Following a JPUC decision to increaserates, both the JTC and a groupof users filed cases

in Jamaica'sSupreme Court. The formerbecause the rate increase was too low, while the latter

because it was too high. The SupremeCourtdeclinedto heareithercase. In 1974, the government

amendedthe JPUC Act to specify a minimumrate of return(relatedto the yield of the Jamaican

goverment's foreign debt) for the companies regulatedunder the Act. The JPUC nevertheless

interpretedthat amendmentas not specifying a "total entitlement"but rathera maximum (see

Spiller and Sampson, 1993).

74. For example, in a 1973 rate case the JPUC rejected the company's request for a price

increase. The governmentthen imposed a tax on telecommunications providedthe company

and

with a subsidy equal to that tax. The JPUC claimed that such action by the government was

illegal, but there was no legal challenge to the government's action. In considerationof the

subsidy scheme, the governmentreceived a 10 percent share in the JTC (Spiller and Sampson,

1993).

75. This section is based mostly on Galal (1993).

234 The Joural of Law., V10

Economics,& Organization, N2





remainedin effect until the governmentinterventionin 1971. The 1930 con-

tract provided CTC with a 50-year concession, indexed its tariffs to "peso-

gold," stipulateda returnon assets up to 10 percent,76 protectedthe company

against terminationof the concession,77 and requiredfrom the company a

particularinvestmentprogramover the next few years. The contractprovided

the governmentwith the ability to take over the company's operationsunder

vaguely defined terms (Galal, 1993). While the 1930 law seems to have

provided an impetus to telecommunications investments, investmentsstarted

to lag and conflict with the governmentover tariffsand costs developed. The

subagreementsentered into in 1958 and 1967 resolved those conflicts. The

1958 subagreementassuredCTC of a 10 percentrate of return(ratherthanup

to 10 percent as in the 1930 law), and CTC agreed to start an eight-year

developmentplan. The 1967 subagreementmaintainedthe rate of returnand

redefinedcosts, and CTC agreedto an expansionplan up to 1971, committing

to increase lines from 1971 on at a rate of 6-7 percentper annum.

Although many other reasons may lie behind the low investment levels

priorto the subagreement 1958,78the vagueness of Chile's telecommunica-

of

tions law did not provide strongsafeguardsagainstopportunistic behaviorby

as

the government,particularly the post-warperiod was one of rapidpolitical

change. The telecommunications statutespecified a maximum,but not neces-

sarily binding, rate of returnof 10 percent.The vagueness of the law had two

main implications:first, regulatorshad substantialdiscretionin its interpreta-

tion; second, because of its vagueness, governmentcould modify the law

throughregulatorydecrees, as it did in 1958 and 1967.79Thus, not only did

the regulatoryregime provide wide latitudeto the regulatorbut therewere no

importantlimitationsto changing the regulatorysystem itself. The subagree-

ments of 1958 and 1967 did not change the regulatoryincentive structure,but

ratherimprovedupon regulatorygovernance.The change in performance that

followed can be related directly to such an improvement.This period, then,

provides an observationfor the Figure 1 branchthat answers no to the "spe-

cific process written in law or contract"question. This period resembles

Jamaica'sJPUC period. In both instances, regulatorygovernancestructures





76. Any return in excess of 10 percent would be divided equally between CTC and the

government.

77. At terminationof the concession the government could either take over the company,

paying the value of its assets expressed in "pesos-gold," or could renew the concession for

another30 years.

78. In particular, refer here to potentialmacroeconomicconsiderationsarisingfrom World

we

War II and its aftermath.

79. In Latin America's presidentialsystems, laws may be interpretedby the administration

through the issuing of regulatory decrees (decretos reguladores) that have the power of law.

Indeed, some legislative acts have no power until such regulatory decrees are issued. As a

consequence, changes in a regulatorydecree may require specific new legislation, unless the

regulatorydecree itself leaves open the need for its own modification(see Shugart and Carey,

1992).

FoundationsRegulatory

TheInstitutional of 235

Commitment





were not alignedwith the countries'institutional and

endowments neither

restrained

regulators'discretion limitedthe potentialfor changingthe

nor

regulatorysystems.In bothcases, network was

expansion limited.



Periods. While the regulatoryschemes of 1958-

4.2.2 The High-Investment

are

71 and 1987 to the present quitedifferent Table3), bothattempt

(see to

limitregulatory discretion. bothcases, limitsto regulatory

In discretion

are

attainedthrough As

specificlegislation. noted,thesubagreements 1958and

of

the

1967 eliminated vaguenessof the previouslaw by assuring CTCa 10

percentrate of return(while at the same time imposingnew investment

The

obligations). post-1987regulatory regimeis of a different

nature. After

experimenting a variety regulatory

with of schemesas a prelude privatiza-

to

tion, Chile's militarygovernment a

introduced new regulatory thatre-

law

mains in force to this day. The new law providesprecisedetailsfor the

of

implementation rate-of-return of

regulation the noncompetitive market

the

segments.Furthermore, law specifiestheprocessto be usedin determin-

ing which services are noncompetitive, and hence subject to regulation. The

processinvolvesa determination the Antitrust

by Resolutive Commission,

following a referencefrom the telecommunications regulator(the Sub-

for

Secretary Telecommunications).

for

Priceregulation the noncompetitive servicesis designedto providea

to

rateof return a theoretically efficient

best-practice, firm.Thespecificity of

law

the telecommunications is suchthatit specifieshow to compute cost the

of

of capital theputatively firm8O howto compute,

efficient and giventhecost

of capital,the maximum prices(basedon a long-run cost

marginal modelfor

each of the individual servicesat the pointof departure) the regulated

for

services. Note, though,that Chile permitsentryeven into these regulated

marketsegments.81 is

This benchmark recalibrated every five years (with

in

indexation the interim periods)by the regulator consultation the

in with

private The

companies. law alsospecifies explicitarbitration

an in

process the

eventof disagreements-with courtsas finalarbiter.82

the

Becausethe law is so specific,regulatory changes new

require legislation,

as the law's specificity

provides verylittlelatitude alternative

for interpreta-

tions. New legislationwouldhaveto overcomeinstitutional obstacles,like

bicameralism and a divided legislaturewith a minoritypresident.The





the

80. Thelawrequires useof thecapital assetpricingmodelin computing allowed

the price

the

levels. Furthermore, law alsostipulates preciseaveraging

the to

procedure be usedin recal-

the

culating costs of the putatively efficientfirm.

81. Chileimposes benchmarkregulation in thosemarket

only that

segments aredeemed to not

and

be contestable hasanexplicit of

procedure public for

hearings determining whethercontest-

abilityexistsin individual

segments Galal,1993).

(see

82. Several aspectsof theregulatory regime to

remained be interpreted theantitrust

by com-

and

missions thecourts-for example, interconnection and

agreements theright Entel(a long-

of

distance company in

created thelate1960sandprivatized in thelate1980s) CTCto enter

also and

intoeachother'sline of business Galal,1993).

(see

236 The Journal Law.Economics,& Organization, N2

of V10





post-1987 period, however, provides a particular source of flexibility that did

not exist in the 1958-74 period. Because regulationis to be undertaken only

in noncompetitive segments, the boundaries of price competition are not

fixed in the legislation but ratherare allowed to be changed by decisions of

the antitrustcommission. It was possible to pass such specific legislation in

the first place, because at the time Chile's governmentwas unified underthe

military,partiallyresolving the legislative obstacles to the creationof a com-

plex regulatorysystem. However, this does not mean that only throughmili-

tary regimes can such specific legislation be achieved. In fact, other demo-

craticallyelected governmentshave introduced similarlyspecific legislationin

other regulatoryareas.83

Thus, during the post-1987 period, and to some extent following the sub-

agreementsin 1958 and 1967, governancestructureswere designed to limit

regulatorydiscretion. These governance structureswere particularlyaligned

with Chile's institutionalenvironment.Specific regulatorylaws in Chile pro-

vide regulatorycredibilitybecause specific legislative changes are not easy to

implement. Finally, deviationsfrom both specific regulatoryinstructionsand

attempts to change the system by governmentalactions can be easily chal-

lenged in the courts. The Chilean judiciary has a record of hearing-and

resolving impartially-regulatory disputes. Indeed, since privatizationit has

become involved in a series of disputessurrounding regulationof telecom-

the

munications concerning, in particular, the determination of competitive

boundaries(see Galal, 1993).

The two periods, however, providevery differentlimitationson regulatory

discretion. While regulatorydiscretionduringthe 1958-71 period was more

limited than duringthe periodpriorto the subagreements 1958 and 1967, it

of

was still much higher thanduringthe post-1987 period. Since 1987, short-run

regulatorydiscretionin pricinghas been almost completely eliminated,as the

1987 law specifies not only the proceduresto be followed, but also the way

informationis to be processed. These episodes, then, provideobservationsfor

the Figure 1 branchthat answersyes to the "specificprocess writtenin law or

contract"question.

We should note, however, that such legislative specificity would not have

provided the same type of commitmentin Jamaicaor the U.K., where laws

can be amendedmuch more easily than in a presidentialsystem like Chile's.

While specificity of the sort designed in Chile could be stipulatedin a license

(a la U.K. or Jamaica), such a high degree of specificity and complexity

requires that those active in enforcing the license-that is, company execu-

tives, bureaucrats,and the courts84-be able to effectively navigateits com-

plexity. While the U.K. bureaucracy traditionallyattractedhighly quali-

has





83. For example, in late 1992 Argentinaintroduceda similar regulatoryregime for the elec-

tricity sector (see Spiller, 1993).

84. We could even include the media, as license renewalshave always been very public affairs

in both Jamaicaand the U.K.

Foundations Regulatory

The Institutional of Commitment 237





it

fledprofessionals, is questionable the

whether Jamaica has

bureaucracy the

necessary depth it

andexperience. a consequence, is reasonable specu-

As to

of

late that the introduction Chilean-style regulatory incentivesin Jamaica

could have triggered continuous litigation,forcing the Jamaican courtsto

makehighlytechnical interpretations, withanunstable outcome beinghighly

probable. termsof Figure1, the U.K. andChileprovide

In observationsfor

the branchthat answersyes to the "strongbureaucracy" question,while

Jamaica provides for that no

observations thebranch answers to thatquestion.

To summarize, evidencepresented

the here is consistent with our main

is to

hypothesis performance related andforemost theextentto which

that first

regulatory governance structures alignedwith the institutional

are endow-

ments of the country.While the incentivestructures Chile'spost-1987

of

periodare particularly strong,theywouldnot be credible the absenceof

in

appropriate regulatorygovernance.



4.3 Rent-Seeking PresidentialSystems: Argentinaand the Philippines

Table4 providesa summary of

chronology telecommunications ownership,

regulation, performance Argentina the Philippines.85

and in and Threefea-

turesareparticularly and

noteworthy will be examined somedetailbelow:

in

first,in neithercountry a workable

has governance

regulatory structurebeen

successfullyput in place; second, telecommunications development in the

Philippines been characterized a "political

has by investment cycle";third,

investment in

levels havebeenreasonable Argentina sinceprivatization, but

In

wereaccompanied veryhighratesof return. thissectionwe attempt

by to

explainandinterpret in

thesepatterns the lightof the framework developed

above.Themainthrust thissectionis as follows:thegapsin theexogenous

of

institutionalendowment the two countries

of for

account theirchronic failure

to establish workable regulatory governance structures;as a consequence,

investmentdecisionshavebeenmadewithveryshorthorizons,

withnegative

for long-term

implications performance.

4.3.1 Regulatory Governance Problems. In 1990 Argentina's main state-

ownedtelecommunications company, ENTel,was splitandwas latersold to

two separateprivateconsortia-one headedby Telef6nicaof Spain, the

secondheaded France

by CableandRadioandStetof Italy.86 rulesunder

The

whichthe private companies wereto operate haverepeatedly

beenchanged,

with pricinga vivid example: set of pricingruleswas announced

one when

privateinvestorswere invitedto bid for ENTel;these ruleswere changed

during with

negotiations the bidders, werechanged twicemorein 1991when

the initialagreements came into conflictwith broader

macroeconomic poli-

cies, andwererenegotiated againin late 1992. Underlying fluidstate

yet the

of the regulatory the

systemaround timeof theprivatization, initialprice

the



85. This section is based mostly on Hill and Abdala (1993) and Esfahani(1993).

86. These companies became known as Telef6nica and Telecom, respectively.

of

238 The Journal Lavw, V10

Economics.& Organization, N2





level for the basic telecommunication pulse87was negotiatedafter the compa-

nies submitted their bids and after the bids were adjudicatedbut before the

companies signed their license agreementsin 1991.88 The two winning con-

sortia negotiatedhard to set the initial price level before taking control over,

and paying for, the two operatingcompanies, suggesting that they were not

confident that the proposed regulatory scheme would actually be imple-

mented. Indeed, although the regulatorysystems specified in the operating

licenses granted at privatizationincorporatedsubstantiverestraintsof some

type or another,what actuallywas implemented(i.e., a price freeze, followed

by indexationto the U.S. consumerprice index) bore no resemblanceto what

was specified in the original licenses.89 We have found no evidence that the

affected companies tried to force implementationof their license specifica-

tions throughappealsto the courts. However, this is not surprising,given the

politicization and corruptionof the Argentinianjudiciary.

The Philippines also has not had a system that imposed any substantive

restraintson regulatorydiscretion. Formally,the telecommunicationssector

has been under the control of a regulatorycommission since colonial times:

before 1972, the Public Service Commission (PSC) was responsible for all

utilities;subsequently,responsibilitiespassed to a specializedtelecommunica-

tions regulatoryagency. However, at no time have either the boundariesof

authorityof the regulatoryagency or the substanceof its regulatorymandate

ever been clearly delineated. Other than the splitting of the PSC into its

componentparts, therehave been no systematiceffortsat telecommunications

regulatoryreform, at least since the Philippinesobtainedits independencein

of

1946. Given the fragmentation the Philippines'legislature, and the poten-

tial for judicial corruption,the system limits the ability of the regulatorto

promotecompetitionwithoutlegislative support.While the traditionally frag-

mentedlegislaturesof Argentinaand the Philippinescould providea basis for

stable regulatoryregimes, in fact both polities' lack of informal restraints,

weak judiciaries, and disregardfor the rule of law have meantthatthe oppor-

tunities for enacting restraintson changing the regulatorysystem are very

limited.90Furthermore, existence of corruptjudiciaries in both countries

the





87. All call prices are based on the price of a "pulse,"with more expensive calls (e.g., long

distance) incurringmore pulses per minute.

88. A similar disdain for rules is reflected in that the exclusive licenses included the service

areas of the main private regional telephone company, CAT, a subsidiary of Siemens. As a

consequence, the licenses took away CAT's legal standing, forcing it to be sold to the new

licensees (see Hill and Abdala, 1993).

89. For example, the original licenses specified a two-yearperiodduringwhich pricingwas to

be set on a rate-of-returnbasis (the rate of returnset at 16 percent, although it was unclear

whetherthis was a maximumor a minimum,with legislatorsand the presidentof ENTeldisagree-

ing on that point), followed by a price-capperiod with X set to zero.

90. Given that the currentregulatoryregime has few substantiverestraints,and that there are

no restraintsto changing the regulatorysystem, it is only a theoreticalexercise to think aboutthe

enforcementof restraintsin the cases of Argentinaand the Philippines.We will not considerthese

issues here, except to note that while the judiciary in the Philippines has played some role in

The Institutional of

Foundations Regulatory

Commitment 239





seems to constitute a basic institutionalflaw, as they cannot be trusted to

restrainthe regulatorsnor to restrainchanges in the regulatorysystems.



4.3.2 Investingin the Face of Adversity. It is reasonableto ask how invest-

ment takes place at all in countrieswith such weak governancestructures.The

answer resides in the extractionof short-termrents. Measuringprofitability

rates in the Philippinesis an impossible task given the extent of corruptionin

procurement can

(Esfahani, 1993), but "short-termism" be seen in the rhythm

of telecommunicationsinvestmentin the Philippines-which is attunedto the

ebb and flow of political actors "friendly"to the telecommunicationsutility

rather than to the business cycle. In particular,Esfahani (1993) shows a

repeated three-pronged investment pattern: first, investment in telecom-

municationstakes off immediatelyfollowing the inauguration governments

of

aligned the

with the groupcontrolling main telecommunications utility(PLDT);

second, investmenttails off in the later years of the regime; and third, invest-

ment stagnates in periods when relations with the group in power are more

distant. As a consequence, sector performancehas been poor, especially in

the provision of local service. Between 1950 and 1991, the numberof tele-

phones in service expandedat the modestaveragerateof 4 percentper annum,

with bursts of moderate growth in fixed assets punctuatedby periods of

stagnation. As of 1991, recordedunmet demandfor telephones amountedto

almost 65 percent of the numberin service.

In Argentina, short-termism apparentin the extremely high profitability

is

of the licensees. For example, for the 11 months ending on September30,

1991, the rate of returnto Telecom's operatorwas 26.9 percent, while that of

Telef6nica's was 203 percent. The returnsto both consortia were also quite

remarkable: 58 percent and 72 percent, respectively (Hill and Abdala,

1993:Table 8). While 1991 investments were not higher than anticipated,

during 1992 both companies increased their number of lines by twice the

number required by their licenses to maintain their 10-year exclusivity

period.91



5. Final Comments and Open Questions

The evidence presentedin the previous section can be interpreted follows.

as

First, privateutilities were aggressive investorswheneverthe threerestraining

mechanisms identified in Tables 2-4 were in place (in Chile, 1958-70 and

post-1987; in Jamaica,pre-1962 and post-1987; in the U.K., post-1984). By

contrast,out of the remainingfour cases wherethe threemechanismswere not





limiting regulatory discretion, this restrainingrole seems to have been tainted by corruption

allegations (see Esfahani, 1993).

91. To a large extent, given the political impossibility of repatriating the cash flow gener-

all

ated by these companies, the fact that they quadrupled their additionsto main lines in 1992 may

simply reflect the use of available cash flow. By reinvesting their retained earnings now, the

companies relax the constraintsthat license investmentrequirements may impose in futureyears,

when cash flow reductionsmay requirethe companies to obtain fresh funds.

240 The Joural of Law,Economics.& Organization, N2

V10





in place, only Argentinahas experiencedany significantprivateinvestmentby

telecommunicationsutilities (althoughonly two years have elapsed since pri-

vatization). And that experience has been accompaniedby unusually high

rates of return.

Second, there were substantialvariationsin the specific forms of the three

mechanisms.These variationsappearto derive from the natureof each coun-

try's exogenous institutions.Chile incorporated substantiverestraintsinto its

regulatory incentive structuresby specifying precisely how regulatedprices

are to be determined.Jamaicaand the U.K. limited administrative discretion

by granting the regulated company freedom to set prices subject to some

overall price constraints(rateof return,in the case of Jamaica,price cap in the

case of the U.K.) and by limiting the ability of the regulatorto interferewith

such decisions. Restraintson changing the regulatorysystem were also in

place in those threecountriesduringthe periodswhen the regulatoryincentive

structureslimited administrative discretion. The characterof these restraints,

however, varied across the threecountries, with the variationsconsistentwith

the countries' exogenous institutionalendowments: the U.K. and Jamaica

used licenses, while Chile used specific regulatorylegislation.

Third, all three countrieswhose regulatorysystems have successfully con-

strained the discretionarypower of regulators have independent and well

regarded judiciaries. And in all threecountries,these judiciarieshave a record

of hearingregulatorydisputes and resolving them impartially. Thus, while in

seven of the nine regulatoryepisodes discussedhere, countrieswere endowed

with exogenous institutions capable of restrainingarbitraryadministrative

action, in only five of these seven episodes did governmentsuse these exog-

enous endowmentsto put in place regulatorysystems that restrainedarbitrary

administrativeaction, and, in turn, were successful in attractingprivate in-

vestment. The remaining two episodes-Jamaica between 1962 and 1975,

of

and Chile before the subagreements 1958 and 1967-appear to be cases of

missed opportunities.In both episodes there was a basic flaw in the design of

regulatory governance-a failure to build substantive regulatoryrestraints

into the system itself. Chile's 1930 law imposed a ceiling (but no floor, until

amended in 1958) on rate of return, and gave the governmentthe right to

intervenein the company's operationsunder vaguely defined circumstances.

Jamaica'sregulatorysystem between 1966 and 1975 was modeledon the U.S.

system and promotedparticipationin an open-endedregulatoryprocess by a

wide range of interest groups-but without the proceduralandjudicial safe-

guards that traditionallyhave protectedutilities in the U.S. Consequently,in

both episodes private utilities eventually failed to invest, and the resulting

conflicts with governmentculminatedin nationalization.

The two remainingcases, Philippinesand Argentina,are more of a mixed

bag. In the Philippines the exogenous domestic institutionshistoricallyhave

provided an inadequatefoundationupon which to erect a regulatorysystem

capable of restrainingadministrativediscretion. Private ownershipseems to

be based on rents extractedthroughthe political process. For all their histori-

cal weaknesses, Argentina'spolitical institutionsmay provide some basis for

The Institutional Foundations of Regulatory Commitment 241







makingcrediblecommitments,as long as thejudiciaryachieves a modicumof

independentcredibility and enforcementcapability.If democracybecomes a

permanentfeature of Argentina, then power is likely to be more fragmented

than it is at present, both between the executive and the legislatureand within

the legislatureitself. Thus, regulatoryreformsthatlimit administrative discre-

tion, either through licenses with very specific and limiting provisions or

through very specific legislation, may prove difficult to change and may

therebyprovide investors with safeguardsfor futureinvestments.The Philip-

pines' stable political institutions, however, seem to frustrateeven such a

mildly positive assessment (Esfahani, 1993).

Our analysis suggests that the foundationof a successful regulatorypolicy

consists of the developmentof a regulatorygovernancestructurethat is ade-

quate, given the nature of the country's institutions, to constrain arbitrary

administrativeaction and that induces private investment to take place. An

exclusive focus on regulatorygovernance, however, is inadequate,as it offers

only limited guidance as to what should be the specific contentof substantive

regulatoryrules. Thus, a unified approachto regulatorypolicy must incorpo-

rate regulatoryincentives (that is, rules concerningpricing, entry, and inter-

connection) into the analysis as well as consider the impact of the specific

content of regulatoryrules on the efficiency with which privateutilities per-

form. Exclusive emphasis on the latter, however, may result in a totally

inadequateregulatorystructure.

At this point it may be useful to speculateon what alternativesare available

to countries that lack the crucial exogenous formal and informalinstitutions

discussed here. Our discussion suggests that in those countriesprivateinvest-

ment will requirethe developmentof alternative safeguards.One exampleof a

safeguardingmechanismis a privatization programthatdistributesshareown-

ership (and thus a stake in the performanceof the privatizedcompany)among

a broad part of the population. Building a broad base of shareholderswas

importantin the privatizationof telecommunications the U.K., and played

in

a modest role in the Chilean, Argentinean,and Jamaicantelecommunications

privatizations(but was an important componentof otherChileanutilityprivat-

izations; see Spiller, 1993). Attempts at widespread ownership require the

prior developmentof a stock market, with relatively well developed security

regulations, which may be lacking in some countries. Similarly, widespread

ownershipmay requirethe developmentof privateinstitutionalinvestors(e.g.

pension funds, insurancecompanies)thatprovidea low-cost conduitfor wide-

spreadand diversifiedstock ownership.The U.K., Chile, and to some extent

Jamaicahave developed these types of institutions,thus facilitatingthe further

development of investment safeguards.92A second option is to privatize

enterprisessequentially (and to have sequential sales of shares in individual

enterprises). Since the success of the later steps of a sequenced program



92. For example, even when the U.K. Labour Party denounced BT's privatizationin the

mid-1980s, its platform called for renationalizationwithout "speculative gains," rather than

outrightnationalization.

Economics, Organization. N2

242 TheJournal Law,

of & V10





depends upon whether the privatizinggovernmentabides by the agreements

made in the earlier steps, the costs to governmentof reneging on its early

agreements can be high. Among the cases studied, Argentinaprovides the

clearest example of this approach. Its privatizationof telecommunications

was the first dramaticstep in a sweeping programto privatizepublic enter-

prises. The potential impact on the remainderof the programafforded the

private buyers of the telecommunicationsutility some confidence that the

Argentineangovernmentwould refrainfromex post administrative expropria-

tion (Hill and Abdala, 1993). Unless the requiredinstitutionsdevelop as the

privatizationprocess progresses, investors will be increasingly reluctantto

invest, because of fears that the end of the privatizationperiod may unravel

the goverment's self-restraint.

As for internationalsubstitutesfor missing nationalfoundations, Jamaica

and the Philippines in the 1950s come closest of all the countriesstudied to

using this mechanism. Jamaica'sjudicial system continues to recognize the

Privy Council in London as the final arbiterof Jamaicancourt decisions, a

feature that may partiallyaccount for its continuedcredibility.Even though

the Philippineswas formally grantedindependencefrom the United States in

1946, for the subsequent 15 years the continuityof pre-independence institu-

tions, the strong leverage of the United States, and specific agreementsthat

protected U.S. investors provided a predictableand safe environmentthat

facilitated investment by both Filipino and U.S. investors.

The potentialexists to go much furtherin using international institutionsas

substitutesfor weaknesses in domestic commitmentcapability.One innova-

tion that has begun to receive attentionis for an international institutionlike

the World Bank to provide private investors (and lenders) with guarantees

against noncommercialrisk, including the risk of administrativeexpropria-

tion. These guaranteesare provided at the request of the host countryof the

investment. In the event of privateinvestorscalling in the guarantee,the host

country becomes liable to repay the international institutionthe value of the

guarantee. A failure to repay would provokea costly ruptureof the country's

relationshipwith an importantinternational institution.Throughsuch guaran-

tees, the country's good standing in the internationalcommunity and its

continuing commitment to regulatory restraint are held hostage to each

other-providing some commitment against administrativeexpropriation.93

In sum, the success of a regulatorysystem dependson how well it fits with

a country'sprevailinginstitutions.If a countrylacks the requisiteinstitutions

or erects a regulatorysystem that is incompatiblewith its institutional endow-

ment, efforts at privatizationmay end in disappointment,recrimination,and

the resurgenceof demandsfor renationalization.







93. Note that a programalong these lines is different from existing programsof insurance

against noncommercialrisk-which shift the costs of administrative to

expropriation the insurer,

and thus do little to enhance the incentive of the host countryto abide by its commitments.

The Institutional Foundations of Regulatory Commitment 243







References

Abueva, Josd V. 1988. "PhilippineIdeologies and NationalDevelopment,"in R. P. DeGuzmin

and

and M. A. Reforma, Government Politics of the Philippines. New York:OxfordUniversity

Press.

Baldwin, Robert, and ChristopherMcCrudden. 1987. Regulation and Public Law. London:

Weidenfeld& Nicholson.

Baron, David P. 1992. Business and Its Environment.Englewood Cliffs, N.J.: Prentice Hall.

Baron, David P., and Roger Myerson. 1982. "Regulatinga Monopolistwith UnknownCosts," 50

Econometrica 911-30.

Barzel, Yoram. 1989. Economic Analysis of PropertyRights. Cambridge,England:Cambridge

University Press.

- . 1982. "MeasurementCost and the Organizationof Markets,"Journal of Law and

Economics 27-48.

Cabral, Luis, and Michael H. Riordan. 1989. "Incentivesfor Cost Reductionunder Price Cap

Regulation,"Journal of RegulatoryEconomics.

Cain, Bruce, John Ferejohn, and Morris Fiorina. 1987. The Personal Vote. Cambridge,Mass.:

HarvardUniversity Press.

Calvert, Randall. 1992. "A Rational Choice Theory of Social Institutions:Cooperation,Coor-

dination and Commitment,"mimeo, University of Rochester.

Cave, Martin. 1991. "Recent Developments in the Regulationof FormerNationalized Indus-

tries," mimeo.

Coase, Ronald H. 1960. "The Natureof Social Cost," 3 Journal of Lawv Economics 1-44.

and

. 1988. "R. H. Coase Lectures,"4 Journal of Law, Economics, & Organization 1-48.

Cox, Gary. 1987. The Efficient Secret. New York:CambridgeUniversity Press.

. 1993. "The Development of Collective Responsibilityin the United Kingdom," 13(1)

ParliamentaryHistory 32-47.

Dimock, MarshallE. 1933. British Public Utilities and National Development.London:George

Allen & Urwin, Ltd.

Duverger,Maurice. 1954. Political Parties:Their OrganizationandActivity in the ModernState.

New York:Wiley.

Esfahani, Hadi S. 1993. Regulations, Institutions and Economic Performance:The Political

Economyof the Philippines'Telecommunications Sector. Washington,D.C.: The WorldBank.

Fesler, James WV. 1983. "The Higher Civil Service in Europe and the United States," in B. L.

Smith, ed., The Higher Civil Service of Europe and Canada: Lessonsfor the United States.

Washington,D.C.: The Brookings Institution.

Fiorina, Morris. 1983. "LegislativeChoice of RegulatoryForums:Legal Process or Administra-

tive Process?"39 Public Choice 33-66.

Galal, Ahmed. 1993. "Regulation, Commitmentand Development of Telecommunicationsin

Chile." The WorldBank.

Galal, Ahmed, Leroy P. Jones, PankajTandon,and Ingo Vogelsang. 1993. Welfare Consequences

of Selling Public Enterprises.NWashington, D.C.: WorldBank. Forthcoming.

Gely, Rafael, and Pablo T. Spiller. 1990. "A RationalChoice Theoryof SupremeCourtStatutory

Decisions, with Applications to the State Farm and Grove City Cases," 6 Journal of Law,

Economics, & Organization263-301.

Goldberg, Victor. 1976. "Regulationand AdministeredContracts,"7 Bell Journal of Economics

426-52.

Hill, Alice, and Manuel Abdala. 1993. Regulation,Institutionsand Commitment: Privatization

and Regulationin ArgentineTelecommunications Sector. Washington,D.C.: The WorldBank.

Hormell, OrrenC. 1928. "Electricityin GreatBritain:A Study in Administration," 27(6) Nation-

al Municipal Review (June 1928).

Huber, Peter WV., Michael K. Kellogg, and JamesThome. 1992. The Geodesic Netwvork 1993

II:

Reporton Competitionin the TelephoneIndustry,WashingtonD.C.: The Geodesic Company.

Jacobson, Gary C. 1990. The Electoral Origins of Divided GovernmentCompetitionin the U.S.

House Elections, 1946-1988. Boulder, Colo.: Westview Press.

V10

244 The Journalof Law,Economics,& Organization, N2





Joskow,Paul L. 1972. "TheDetermination the Allowed Rate of Returnin a FormalRegulatory

of

Proceeding,"3 Bell Journal of Economics & ManagementScience 632-44.

. 1973. "PricingDecisions of RegulatedFirms:A BehavioralApproach,"4 Bell Journal

of Economics & ManagementScience 118-40.

--- . 1974. "Inflationand Environmental Changein the Process of Public

Concern:Structural

Utility Price Regulation," 17 Journal of Law and Economics 291-328.

- , and Roger G. Noll. 1981. "Regulationin Theory and Practice:An Overview."In G.

Fromm, ed., Studies in Public Regulation. Cambridge,Mass.: MIT Press.

, and Richard Schmalensee. 1983. Marketsfor Power. Cambridge,Mass.: MIT Press.

Keen, FrankN. 1925. The Law Relating to Public Service Undertakings.London:P. S. King &

Sons Ltd.

Kerry,Michael. 1983. "Administrative Law and the Administrator,"

38(3) Managementin Gov-

ernment 168, cited in Baldwin and McCrudden, 1987.

Knieps, Giinter, and Pablo T. Spiller. 1983. "Regulatingby PartialDeregulation:The Case of

Telecommunications," Administrative

35 Law Review 391-422.

Laffont, Jean Jacques, and Jean Tirole. 1986. "Using Cost Observationto Regulate Firms," 94

Journal of Political Economy 614-41.

--- , and -- and

. 1993. A Theoryof Incentivesin Procurement Regulation. Cambridge,

Mass.: MIT Press.

Loeb, Martin, and Wesley Magat. 1979. "A DecentralizedMethod of Utility Regulation,"22

Journal of Law and Economics 399-404.

and

Lupia, A., and MatthewD. McCubbins. 1993. "WhoControls?Information the Structure of

Legislative Decision Making," mimeo, University of California, San Diego.

McCubbins, Mathew D., Roger G. Noll, and BarryR. Weingast. 1987. "Administrative Proce-

dures as Instruments PoliticalControl,"3 Journalof Law,Economics, & Organization243-

of

77.

---

----, and . 1989. "Structure Process, Politics and Policy: Administrative

and

Arrangementsand the Political Controlof Agencies," 75 VirginiaLaw Review 431-82.

Mathios, Alan, and RobertRogers. 1989. "The Impactof AlternativeFormsof State Regulation

of AT&Ton Direct Dial, Long DistanceTelephoneRates."20(3) RANDJournalof Economics

437-53.

Miller, Charles. 1985. Lobbying Government:Understandingand Influencingthe Corridors of

Power. Oxford, England:Basil Blackwell Ltd.

for

Mueller, M. 1993. "Commentson Interconnection European Telecommunications." Presented

at the 21st AnnualTelecommunications Policy ResearchConference, Solomon Islands, Mary-

land, October.

Newman, K. 1986. The Selling of BritislhTelecom. London:Holt, Rinehartand Winston.

North, Douglass C. 1990. Institutions,InstitutionalChange, and EconomicPerformance.Cam-

bridge, England:CambridgeUniversity Press.

?-----. 1981. Structureand Change in Economic History.

A

, and RobertP. Thomas. 1973. The Rise of the WesternWorld: New EconomicHistory.

Cambridge,England:CambridgeUniversity Press.

and

Palmer,M. S. R. 1992. "TheEconomicsof Organization MinisterialResponsibility: Towards

a Frameworkof Analysis for WestminsterGovernment,"mimeo.

Riordan, Michael H. 1992. "Relevanceof the New RegulatoryEconomics for Telecommunica-

tions." The WorldBank.

Salzberg, Eli M. 1991. "The Delegation of Legislative Powers to the Courts and the Indepen-

dence of the Judiciary,"mimeo.

Schwartz, Edward, Pablo T. Spiller, and Santiago Urbiztondo. 1993. "A Positive Theory of

Legislative Intent,"57(1 & 2) Law and Contemporary Problems51-86.

Self, Henry,and ElizabethM. Watson. 1952. ElectricitySupplyin GreatBritain:Its Development

and Organization. London:George Allen and Unwin Ltd.

Shepsle, Kenneth, and Barry Weingast. 1981. "Structure-Induced Equilibriumand Legislative

Choice," 37 Public Choice 503-19.

of Commitment 245

Foundations Regulatory

The Institutional





, and . 1987. "InstitutionalizingMajority Rule: A Social Theory with Policy

Implications,"AEA Papers and Proceedings 367-371.

- , and . 1987. "The InstitutionalFoundationsof CommitteePower," 81 American

Political Science Review 85-105.

, and -- . 1989. "PenultimatePower: Conference Committeesand the Legislative

Process," Hoover InstitutionWorkingPaper #P-89-7, StanfordUniversity.

Shin, Richard T., and John S. Ying. 1992. "UnnaturalMonopolies in Local Telephone,"23

RAND Journal of Economics 171-83.

Shugart, Matthew S., and John M. Carey. 1992. Presidents and Assemblies. New York:Cam-

bridge University Press.

and

Spiller, Pablo T. 1993. "Institutions RegulatoryCommitmentin Utilities' Privatization," 2(3)

Industrialand Corporate Change 387-450.

- , and Rafael Gely. 1992. "Congressional Controlor JudicialInterpretation: Determi-

The

nantsof US SupremeCourtLaborDecisions: 1949-1988," 23(4) RANDJournalof Economics

463-92.

--- , and Brian Levy. 1991. "Regulation, Institutionsand Economic Efficiency: Promoting

Regulatory Reform and Private Sector Participationin Developing Countries;A Research

Proposal," mimeo, University of Illinois.

and The

, and Cezley I. Sampson. 1993. "Regulation,Institutions Commitment: Jamaican

TelecommunicationsSector."The WorldBank.

, and Santiago Urbiztondo. 1993. "Political Appointees vs. Career Civil Servants: A

Multiple PrincipalsTheory of Political Bureaucracies,"forthcomingin EuropeanJournal of

Political Economy.

, and Luis Viana. 1992. "How Not to Do It: ElectricityRegulationin Argentina, Brazil,

Chile, and Uruguay,"mimeo, University of California, Berkeley.

, and Ingo Vogelsang. 1993a. "Regulation, Institutionsand Commitment:The British

TelecommunicationsSector." The WorldBank.

, and . 1993b. "Notes on Public Utility Regulation in the UK: 1850-1950,"

mimeo, University of Illinois.

Spulber, Daniel. 1989. Regulationand Markets. Cambridge,Mass.: MIT Press.

Stone, Carl. 1981. "PartyVoting in Jamaica (1959-1976)," in C. Stone and A. Brown, eds,

Perspectives on Jamaica. Kingston, Jamaica:JamaicaPublishingHouse.

. 1986. Class, State and Democracy in Jamaica. New York:Praeger.

Taagepera,Rein, and MatthewS. Shugart. 1993. "Predictingthe Numberof Parties:A Quantita-

tive Model of Duverger Mechanical Effect," 87 AmericanPolitical Science Review 455-64.

Veljanovski, Cento G. 1991. Regulators and the Market: An Assessment of the Growth of

Regulation in the UK. London:Instituteof Economic Affairs.

WVeingast, BarryR. 1984. "The Congressional-Bureaucratic System: A PrincipalAgent Perspec-

tive," 44 Public Choice 147-92.

. 1993. "The Economic Role of Political Institutions,"mimeo, StanfordUniversity.

, and M. Moran. 1983. "Bureaucratic Discretion or CongressionalControl:Regulatory

Policy Making by the Federal Trade Commission," 91 Journal of Political Economy 765-

800.

Williamson, Oliver E. 1975. Marketsand Hierarchies:Analysis and Antitrust Implications.New

York:Free Press.

-. 1976. "FranchiseBidding for NaturalMonopolies: In General and With Respect to

CATV,"7 Bell Journal of Economics 73-104.

. 1979. "Transaction-Cost Economics: The Governance of ContractualRelations," 22

Journal of Law and Economics 3-61.

~---. 1985. The Economic Institutionsof Capitalism. New York:Free Press.

. 1988. "TheLogic of Economic Organization," Journalof Law, Economics, & Organi-

4

zation 65-93.

. 1991. "Comparative Economic Organization: Alter-

The Analysis of Discrete Structural

natives," 36 Administrative Science Quarterly269-296.

Economics.& Organization, N2

246 The Journalof Lavw. V10





. 1993. "Publicand PrivateBureaus: Transaction

A Costs Economics

Assessment,"

of

mimeo,University California, Berkeley.

WorldBank.1992."Bolivia: for Part

Restructuring Growth, One:TheMainReport," Washing-

ton, D.C.: TheWorld Bank.

Young, Hugo. 1985. "Judgesand the Exercise of Power,"The Guardian, October 31 (p. 31).

Zupan, Mark. 1989. "CableFranchise Firms Behave Oppor-

Renewals:Do Incumbent

tunistically,"20 RAND Journal of Economics 473-82.