Public Disclosure Authorized

Public Disclosure Authorized

Public Disclosure Authorized
Public Disclosure Authorized

LV)?Al2/2

212 EI

World Bank Discussion Papers

Institutional

O)ptions for
the Provision
of Infrastructure
Christine Kessides

Recent

World

Bank

Discussion

Papers

No. 155

Managing External Debt in Developing Counttries: Proceedingsof aJoint Seminar,Jeddah, May 1990. Thomas M.
Klein, editor

No. 156

Developing Agricultural Extensionfor Women Farners. Katrine A. Saito and Daphne Spurling

No. 157

Awakening the Market: Viet Namn'sEcotiomic Transition. D. M. Leipziger

No. 158

Wage Policy during the Transitiotnto a Market Economy. Poland 1990-91. Fabrizio Concelli and Ana Revenga,
editors

No. 159

Intemational Trade and the Environment. Patrick Low, editor

No. 160

Intemational Migration and IntermationalTrade. Sharon Stanton Russell and Michael S. Teitelbaum

No. 161

Civil Service Reform and the World Bank. Barbara Nunberg and John Nellis

No. 162

Rural Enterprise Development in China, 1986-90. Anthony J. Ody

No. 163

The Balance between Publicand Private Sector Activities in the Delivery of Livestock Services. Dina L. Umali, Gershon
Feder, and Comelis de Haan

No. 164

How Do National PoliciesAffect Long-run Growth?: A ResearchAgenda. William Easterly, Robert King, Ross
Levine, and Sergio Rebelo

No. 165

FisheriesDevelopmeit, FisheriesManagement, and Externalities. Richard S. Johnston

No. 166

n e Building Blocks of Participation: Testing Bottom-up Plantning. Michael M. Cernea

No. 167

Seed System Developneti: The Appropriate Roles of thiePrivate and PublicSectors. Steven Jaffee and Jitendra Srivastava

No. 168

Environmental Managemenitand Urban Vulnerability. Alcira Kreimer and Mohan Munasinghe, editors

No. 169

Cotnmotn PropertyResources: A M-fissing

Dimerision of DevelopmenttStrategies. N. S. Jodha

No. 170

A Chinese Provinceas a Reform Experiment: 7he Case of Hainan. Paul M. Cadario, Kazuko Ogawa, and Yin-Kann Wen

No. 171

Issuesfor Itifrastruicture
Management in the 1990s. Arturo Israel

No. 172

Japanese National Railways Privatization Study: TheExperience ofJapan and Lessonsfor Developing Countries.
Koichiro Fukui

No. 173

The Livestock Sector in Eastern Europe: Constraints and Opportunities. Comelis de Haan, Tjaart Schilhom
Veen, and Karen Brooks

No. 174

Assessing Development Finatice Institutionis:A Public Interest Analysis. Jacob Yaron

No. 175

ResourceManagement and Pastoral Institution Building in the West African Sahel. Nadarajah Shanmugaratnam,
Vedeld, Anne Mossige, and Mette Bovin

No. 176

Publicand Private Sector Roles in Agricultural Research: Tlieory and Experience. Dina L. Umali

No. 177

The RegulatoryItnpedimentsto the P-nvateIndustrialSectorDevelopmnent

in Asia: A ComparativeStudy. Deena Khatkhate

No. 178

China: Reforming Intergovemmental

No. 179

Nippon Telegraphand Telephone Privatization Study: Experience ofJapan and Lessonsfor Developing Couintries.
Yoshiro Takano

No. 180

China's Refonn Experience to Date. Peter Harrold

No. 181

Combatting AIDS atid Other Sexually Transmitted Diseases in Africa: A Review of the World Batik's Agendafor Actionl.
Jean-Louis Lamboray and A. Edward Elmendorf

No. 182

Privatization Problems at Industry Level: Road Haulage in Central Europe. Esra Bennathan and Louis S. Thompson

Van

Trond

Fiscal Relations. Ramgopal Agarwala

(Continued

on the inside back cover.)

212~

World Bank Discussion Papers

Institutional
O ptions for
the Provision
of Infrastructure
Christine Kessides

The World Bank

Washington, D.C.

Copyright C 1993
The International Bank for Reconstruction
and Development/T lHE WORLD BANK
1818 H Street, N.W.
Washington, D.C. 20433, U.S.A.
All rights reserved
Manufactured in the United States of America
First printing September 1993
Second printing February 1995
Discussion Papers present results of country analysis or research that is circulated to encourage discussion
and conunent wvithin the development community. To present these results with the least possible delay, the
typescript of this paper has not been prepared in accordance with the procedures appropriate to fomial
printed texts, and the World Bank accepts no responsibility for errors. Some sources cited in this paper may
be infonral documents that are not readily available.
The findings, interpretations, and conclusions e.xpressed in this paper are entirely those of the author(s) and
should not be attributed in any manniier to the World Bank, to its affiliated organizations. or to members of
its Board of Executive Directors or the countries they represent. The World Bank does not guarantee the
accuracy of the data included in this publication and accepts no responsibility whatsoever for any
consequence of their use. The boundaries, colors, denominations, and other infonnation shown on any map
in this volume do not imiply on the part of the World Bank Group any judgmenit on the legal status of any
territory or the endorsement or acceptance of such bounldaries.
The material in this publicationi is copyrighted. Requests for permission to reproduce portions of it should
be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank
encourages dissemniationi of its work and wvillnormally give permission promptly and, when the
reproduction is for nonconunercial purposes, without asking a fee. Permission to copy portions for classroom
use is granted through the Copyright Clearance Center, Inc., Suite 910, 222 Rosewood Drive, Danvers,
Massachusetts 01923, U.S.A.
The complete backlist of publications from the World Bank is shown in the anniual Inidex of Publicatiotis,
which contains an alphabetical title list (with full ordering infomiation) and indexes of subjects, authors, and
countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the
Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from PLublications,
The World Bank, 66, avenue d'Ina, 75116 Paris, France.
ISSN: 0259-210X
Christine Kessides is a senior economist in the Urban Development Division of the World Bank's
Transportation, Water and Urban Developmelnt Department.
Library of Congress Cataloging-in-Publication

Data

Kessides, Christine, 1953Institutional options for the provision of infrastructure /

Christine Kessides.
p.
cm. - (World Bank discussion papers, ISSN 0259-21OX
212)
Includes bibliographical references.
ISBN 0-8213-2627-9
1. Infrastructure (Econormucs)-Finance.
2. Capital investments.
3. Economic assistance, Domestic.
1. Title.
11. Series.
HC79.C3K45
1993
350.86-dc20
93-27361
CIP

iii
ABSTRACT
This paper developsa comprehensiveframeworkfor identifyingthe criteria for choosingamong
various institutionalforms for infrastructureactivities.The frameworkis based on the presumptionthat
the objectivesof the decision are to promote efficiency, fairness, and accountabilityin the supply of
infrastructureservices. The discussionstarts from the premise that the competitiveprivate market is the
preferred modeof supplywhen the economicand technologicalcharacteristics of the activity permit
it. The characteristicsof various infrastructureactivitiesare discussedto identify the scope for applying
competitivemarkets, and the appropriatefunctionsto be played by the governmentand private sector.
The intentionhere is to make the rationalefor choice as transparentas possible. However, this process
of analysisdoes not lead to one unique institutionalsolutionfor any infrastructureactivityin all contexts.
A variety of institutionalarrangementswhich involvedifferentdegreesof involvementby the public and
private sectors are discussedwith referenceto specificexamplesin developedand developingcountries.
The paper then summarizeskey policy issuesthat affect the actual incentivesand performanceof these
institutionalarrangements-includingissuesof creatingcompetition,broadeningparticipation,achieving
appropriate treatment of risks, and designing effective regulation, financing, and planning of
infrastructure.

iv

ACKNOWLEDGEMENTS
The initialconceptfor Part I of this paper was drawn from a note by Rita Hilton. She also made
significantcontributionsto the design of the subsectoraltables (AppendixU). ThelmaTriche provided
manythoughtfulinsightsto the discussionof contractingarrangementsin Part Two. Carl Bartone, Peter
Cordukes,Harvey A. Gamn,Timothy Nulty, Herve Plusquellec,Louis Thompson,and GuillermoYepes
provided inputs to the developmentof the subsectoraltables. Thanks are also due to the following
reviewerswho commentedon various drafts of this paper: Patricia Annez, Eliot Berg, MichaelCohen,
James F. Ford, Alice Galenson, Pierre Guislain, KennethGwilliam,Frannie Humplick, Arturo Israel,
loannisKessides,CarolineMoser,Louis Pouliquen,and Louis Thompson.Kari Labrie and Luisa Sambeli
processed the document.Any errors that remain are the author's own.

ABBREVIATIONS
BOT
BOO
MSW
NGO
O&M
UFW
WSS

Build-Operate-Transfer
Build-Own-Operate
MunicipalSolid Waste
Non-governmentalorganizations
Operationsand maintenance
Unaccountedfor water
Water Supply and Sanitation

FOREWORD
Infrastructure has been a major focus of World Bank lending since its founding, and currently
accountsfor abioutforty percent of our portfolio. A key concernof the Bank is to ensure that countries
3bt;.in the ful benefits from their substantial investmentsin nfrastructure. Our review of operational
experienceand research regarding infrastructuredevelopmentpoints to the conclusionthat institutional
issues-espec!ally those issuesconcerningthe relativeroles of governmentand the private sector, and the
strecture of incentives-are the key to performance in these sectors. In the past decade, international
experiencewith reform and restructuringof infrastructureactivitieshas been growingrapidly.This paper
attempts to develop a systematic approach to evaluating the range of institutionaloptions in the
infrastructuresectors; to take stock of the lessons of experienceand identify the best practices; and,
hopefully, to accelerateand focus the learning process.
The present studywas undertakenas part of an or, ing effort in the Transport,Water and Urban
DevelopmentDepartmentto look at the key cross-sector. issuesin infrastructure.This analysissupports
the Bank's internal review of its own policy and operationalactivities in infrastructure,as well as the
Bank's dialoguewith borrowersand other agen.ies on infrastructureissues.The paperdrawsupon much
recent analyticalwork, done within and outside the Bank, as well as on numerouseconomicand sector
studies by Bank staff which evaluate the perfermance and impact of infrastructurein the context of
individual developing countries. The intended audiencefor this paper includes students of economic
developmentas well as officials, in both developedand developingcountries, who are concernedwith
infrastructurepolicy.

vii
CONTENTS
EXECUTIVE SUMNIARY .

....................................... ix

I
INTRODUCTION ............
................................
1. A FRAMEWORK FOR ANALYSIS OF INSTITUTIONAL ARRANGENIENTS
3
IN INFRASTRUCTURE ......................
4
A. Determining the "Marketability" of Infrastructure Activities ..................
4
The Nature of the Good Involved
......................
......................
5
The Nature or Conditionisof Production
7
......................
Externalities alnd Social Objectives
8
Characteristics of Deniuiid and Service Use .................
14
B. Determining the Public and Private Roles in Infrastructure ..................
C. Characteristics of' Alternative Institutional Arrangements ...................
17
11. LESSONS OF EXPERIENCE WITH ALTERNATIVE
ARRANGEMENTS ..............
A. Review of Experience by Type of Arrangement .21
Governienit Depirtmienti ... ......................................
Public Utiliy. .23
Service Contracting..27
Managemenit Contracts .29
Lease Contracrs ..
.............................................
Concession1s........................
...........
Private Entrepreneurship ........................................
B. Comparison and Evaluation of Alternative Institutional Arrangements

INSTITUTIONAL
21...
I
21

30
331
35
36

III. CONDITIONS FOR ENSURING THE EFFECTIVENESS OF INSTITUTIONAL

ARRANGEMENTS
..
39
A. Creating Competition .39
B. Developing Effective Participation .40
C. Reducing Risks
43
D. Effective Regulation-" Doing Well Only What is Necessary".45
E. Pricing and Financing.................
48
50
F. Investment Planning..
ANNEX I - GLOSSARY ........................................

5I

ANNEX
II - CHARACTERISTICS
OF
INFRASTRUCTURE
SECTORS/SUBSECTORS-A BREAKDOWN BY ACTIVITY ....
........
..

53

ANNEX III - PREVALENCE OF CERTAIN FORMNS OF PRIVATE SECTIOR

PARTICIPATION IN INFRASTRUCTURE BY SECTOR
..............
...
69
REFERENCES ..............................................

76

ix
EXECUTIVE SUMMARY
The present study provides both an analyticalframework, and a review and assessmentof
experience, regarding institutional reform issues and options in infrastructure. "Institutions" are
understood here to include factors affectingthe relationshipsamong actors (especiallygovernmentand
private parties) and the structure of incentives;thus, the classic competitivemarket is one type of
institution.The key concern of this paper is to clarify the rationalesfor choosingparticular institutional
arrangementsfor the provisionof infrastructureservices, with the aim of promotingefficiency,equity,
and accountabilityto users and other financiers.These choicesshould be based on an understandingof
the economicand technologicalcharacteristicsof the infrastructureservices; the incentivesprovidedby
various institutionalarrangements;and the issuesinvolvedin implementingthe institutionalarrangements
so as to achievethe desired outcomes.
Much of infrastructurehas characteristicsthat make it unsuitable to be provided solely
throughcompetitivemarkets.Suchfeaturesof "marketfailure" includeeconomicand technicalconditions
which create natural monopoly,"lumpy" investmentrequirements,spillover effects on non-users, and
the fact that some services are consumedjointly as "public goods". These factors create a legitimate
public interest in infrastructure, but they do not necessarily require government involvement in all
aspects of providing the services. Moreover, it has become increasingly recognized that these
characteristicstraditionallyattributedto infrastructureare not equallytrue of all the activities involved.
For example, in railways, only the imbeddedrail infrastructurehas monopolyfeatures, (and even these
maybe very limitedby competitionby othermodes), whileoperationof the railwayequipmentdoes not.
In manyactivities,technologicaldevelopmentshavewidenedthe scope for competition(especiallyin long
distancetelecommunicationsservices);createdlow-cost supplyoptions(e.g. in sanitationand irrigation);
and increased the possibilitiesfor pricing of individualconsumptionand charging for spillover effects
such as pollution (e.g. electronicroad pricing). Technologyhas also created new types of services-for
example, informatics and intermodal transport systems have revolutionizedtrade logistics-and new
options for production.Even the nature of externalitiesand socialconcernsin infrastructurevary greatly
by type of infrastructure:for example, urban roads and solid waste disposal systems have stronger
impactson local communitiesand on land use patterns thando poweror telecommunicationsdistribution
systems. Thus infrastructure is highly heterogeneous, the traditional "monolithic" structures of
supplying infrastructure are no longer relevant in many activities.
The worldwide trend towards economic liberalization in many sectors, including
infrastructure,has led to experimentationwith various institutionalarrangements.This experiencehas led
to two importantconclusions:that there are fewer activities requiring public intervention than once
was believed; and that public intervention, when justified, can be exerted through less distorting
policy instruments than those traditionally used.
The pervasivegovernmentinvolvementin finance, regulation,and actualdelivery of many
infrastructureserviceshas led to poor performancein manycases-by weakeningmanagers' operational
and financial responsibility, imposingconflictingobjectives, and politicizingdecisions on investment,
pricing, labor, and technologicalchoice. It is necessaryto reexaminethe basic rationalefor government
interventionand the justificationfor eachtype of policyresponse.In many cases, experience shows that
the risks of poor performance due to possible market failure are less serious to the economy than
the risks of government failure. Public infrastructuremonopolieswere created in many countries in
order to exploit economiesof scale, protect nationalinterests, and mobilize resources for basic system
development.In practice, manyof these entitiesachieveneitherefficiencynor distributionalfairness, and

x
they fail to generatesustainablesources of finance (whetherfrom internal revenues, loans or equity).
Moreover,the extremelypoor qualityand unreliabilityof servicehandicapseconomicactivitieselsewhere
in the economy.
Clarifythe rationalefor private versus public involvementin infrastructure.The first issue
for analysisis to determinewhether market conditions exist for each infrastructure activity, based
on the followingcriteria:
(a)

nature of the good/service-whether jointly consumed("publicgood") or privately

consumed("private good");

(b)

conditions of production-to what extent there are economies of scale creating

natural monopoly; whether there are high sunk costs which would deter new
suppliers (if not, the activity is said to be 'contestable"); and what degree of
coordination(e.g., of technicalstandards)in productionis neededfor efficiency.

(c)

externalities and social objectives-to what extent are there benefits and costs
affectingpersons other than those directly involvedin the activity;

(d)

characteristics of user demand-such as the degree of consumers' access to

informationabout supply alternatives,and the existenceof substitutesfor particular
kinds of services.

The aboveconditionsprovidethe a priori justificationfor assigningto the private sector or

to the public sectorresponsibilityfor various functions-including sectoral planningand policy-making,
ownership,regulation,financing,executionof investment,and/oroperationand maintenance(O&M)-for
each infrastructureactivity. The followinggeneral guidelinesfor policy are suggestedby this analysis:
*
For activities involvingpublic or quasi-public goods, natural monopoly,or capital
with high sunk costs-e.g., the provision of network, trunk-type facilities such as power transmission
grids, major highwaysand pipelinesfor water and sanitation,port installations,etc.-there is a case for
a public sector role in planning/policy-making,financing and ownership; or alternatively,for private
sector ownershipunder public regulation.

*
The activitiesnecessaryto generateservicesfrom thesefacilitiesmay best be carried
out on the basis of competitivebidding for the right to operate the monopoly. The government's
responsibilityin this case is to issue the exclusivecontractand monitorperformanceunder its terms; to
ensurethat other providersof servicesusingthe networkfacilitiesface fair conditionsof access(including
price); and to protect users from other possible abusesof the monopoly.
*
For most activitiesinvolvingcapitalwith low sunkcosts (e.g. road freighttransport),
entry by the private sector should be fully liberalized(urban bus transport being an exceptionwhere
regulationof entry can still be justified);the governmentshould mainlybe responsiblefor ensuring fair
competition.
*
Additional characteristicsof infrastructure activities such as externalities, social
service objectives,and certain featuresof user demandmay providejustificationfor public intervention

xi
through investmentplanning, regulationand/or fiscal transfers (taxes/subsidies)-but rarely for public
ownershipor direct public executionof investmentor service operation.
Choose institutionalarrangementswith appropriate assignmentof responsibilitiesto the
public and private sectors. The institutionaloptions consideredhere represent a continuumfrom mainly
public sector, to mainly private sector, responsibilitiesfor the functionsof planning/policy-making,
ownership, regulation, financing, investment, and O&M. These options include: (i) government
department, (ii) public utility, (iii) service contracts, (iv) managementcontracts, (v) lease contracts
(affermages),(vi) concessions,includingBOTs, (vii)private entrepreneurship(i.e., with at least majority
private ownership),includingthroughdivestiture,and (viii)communalor "self-help" schemes,including
cooperatives.Much of the prevailingattentionto 'privatization' in infrastructuretends to focus on only
a few of these options (notablyBOTs and divestiture).
The choices amongthese institutionaloptionsfor particular infrastructureactivitiesshould
depend in large part on the objectives,or benefits, sought from involving the private sector. These
objectives include (i) skilled and independentmanagement,(ii) productive efficiency, (iii) innovation
(dynamicefficiency),(iv) accountabilityto customers(and improvedservice quality), and (v) financial
autonomy(includingmobilizationof additionalfinancialresourcesfor the sector).
Experiencewith the alternativeinstitutionalarrangementsin various infrastructureactivities
leads to the followingobservationsregardingtheir potentialadvantagesand disadvantages:
*
Some improvementsin managementand productive efficiencycan be gained from
introducing commercialpractices (such as by shifting activities from a government department to a
corporatizedparastatal)and even a modestdegree of private sector participation(as in contracting-out
of specific services or managementof operations); these approaches are often impeded by political
interference,however,so that the benefitsmaynot be sustained.Such incrementalarrangementsmay still
be useful as part of a transition to fuller private sector involvement.

Arrangementswhich give the private sector producer of services full managerial

autonomy and full commercial risk (e.g., lease contracts), and in addition, responsibility for both
operation and investment (e.g., concessions)have the potential to produce stronger and more lasting
benefitsthan more limitedforms of privatesectorparticipationsuch as servicecontractsand management
contracts.
*

*
BOTs and completeprivate entrepreneurshipthrough divestituremay, in addition,
mobilizenew sourcesof fundingand furtherreduce government'sfinancialrisk-but this dependson the
terms of the specific agreements.
*
Service contractsand managementcontractsare arrangementswhich appear to have
unexploitedpotentialas part of a strategyof transitionto greater privatization.As the public and private
sector gain experiencewith these partnerships,they can lead to progressiveexpansionof private sector
participationthrough leasesand concessionsand in some cases, ultimately,to divestiture.
Create the conditionsfor successfulimplementationof institutionalarrangements.The actual
incentives involved in these institutional arrangements depend on the framework for competition,
participation,regulationand financing.

xii
Competition.Even where legal barriers to entry in certainactivitiesare removed,there are
often practical constraintswhich prevent new entrants from competingon equal terms with existing
suppliers-such as regulationsregarding accessto fixed facilities(e.g., transfer stations, bus terminals)
or to credit, or tax rules that favor public firms. To increase the pool of potentialentrants, it may be
helpful in some cases to develop opportunitiesfor competitionbetweenpublic and private operators, as
is practiced in a numberof countriesfor solid waste collection.To reduce uncertaintyand the scope for
rent-seeking (corruption)by both private and public sector partners in contractualarrangements,it is
necessaryto provide disclosureof information(about the conditionof fixed assets, for example),and a
public process of bidding for contracts.
Participation. Institutionsmust be adaptedto foster participationby both users and other
interest groups, includingthose affectedby environmentalor other externalities.Participationoptions
include the provision of facilities and services through self-help schemes (e.g. by neighborhood
associations),or cooperativesas practicedin manycountries for electrificationand telecommunications
in rural areas; and through mechanismsto represent the views of users and other stake-holders in
decision-makingregarding investmentplanning, regulation,and delivery of services.
Experiencewith participatoryapproachesin infrastructureactivitiessuggeststhat involving
users in all phasesof the project cycle (i.e., in the initialdesign as well as implementation),and providing
opportunitiesnot merely for consultation,but rather for responsibledecision-making,is likely to produce
the greatest benefits in terms of both project effectivenessand building of local capacity for sustained
development.Specificeffortsoften needto be made at the outset to promotethe involvementof the very
poor and representationof womenas well as men in participatoryprograms.
Regulation. To design appropriate regulatory policy in any infrastructureactivity, it is
necessaryto identifyclearly the rationalefor governmentinterventionin each area as discussedabove,
and the specific objectivesto be sought. A basic prerequisitefor regulationis a stable, predictablelegal
frameworkwhich is enforced, especiallyregardingpropertyrights, liability, and contracting.
Regulatoryfunctionsand responsibilitiesshould be clearly separatedfrom those of system
operations;this can be achievedthrough organizationalseparation(e.g. independentregulatory agencies
distinct from the operating entities, as in the U.K. utilities sectors) or by delegationthrough contracts,
as is the practice in French urban services. Regulatoryproceduresshould also be transparent, easy to
administer and enforce promptly. The scope for regulators to exercise discretion should be clearly
circumscribedin most cases, in order to createsufficientconfidencein the stabilityand objectivityof the
process. Where regulationof tariffs is necessary(for monopolyservices),methods should be preferred
which lead to simple, automatic adjustments and enable producers to benefit from efficiency
improvements.It is also necessarythat regulators have direct access to informationabout quality of
service and user satisfaction,with mechanismsfor consultationwith the public.
For governments unaccustomedto formal regulation, the practice of contracting for
operations(throughservicecontracts, managementcontracts,and leases)mayprovide a gradual learning
process, with the regulatory functionembodiedin the design and monitoringof the contract.This may
be a morepracticalapproachin some developingcountriesthan creatinga separatenew regulatoryagency
at the outset, especiallyfor subsectorssuch as railways,urban water supply and sanitation.
Pricing and Financing.Infrastructureis more likely to be economicallyefficient, and to
have favorable impacts on the environment,when it is subject to user charges. The absence of user

xiii
chargeshas usuallynot promoted accessto servicesby the poor, but has rather reduced availabilityand
worsenedinequalities.User charges shouldbe based on economicprices reflectingboth costs of supply
and demand considerations(willingnessto pay) and, to the extent possible, externalities. In many
infrastructure activities, the structure of tariffs needs to be revised to remove or reduce internal
cross-subsidies, such as between long distance and local telephoneservice, or between industrial and
householdusers. As institutionalreformspermit facilitiesand servicesto be "unbundled"(managedunder
separateorganizationalstructures)and increasedentry by the private sector is permitted,the structureof
rates becomes even more important, as it determinesthe incentivesfor investmentand operation in
different market segments.
Social objectives (for legitimatepublic service obligations, such as supply of water and
transport to low-income neighborhoods)should be financed by explicit budgetary subsidies to the
providerfor these purposesalone; or, an evenbetter approachwhere feasible, paid directly to the needy
population.Suchtransferpaymentsmust be clearlyspecifiedat the outset in any contractingarrangement
with private providers. Moreover, they shouldpreferablybe auctioned-offto bidders who will minimize
the costs of such services. Beforeany such subsidiesarejustified, the ability and willingnessto pay for
servicesshould be clearly assessed.
In activitieswhere servicescannotbe pricedaccordingto individualconsumption,financing
through benefit taxes (i.e. paid by the beneficiariesof the services)provides an instrumentto manage
demand and promote efficientresource allocation. Establishingan explicit link between such revenues
and the activities they support can be an important element in transforming activities such as road
maintenanceand municipalsanitationfrom a bureaucraticto a commercialorientation.
Once incentivesfor internalcost recoveryand financialdisciplineare clearlyestablishedin
the infrastructuresectors, the prospectsfor accessto externalfinancingimprove.Debt instrumentsshould
be used in many areas of infrastructure,especiallyin the creation of long-lived assets to ease the cash
flow and promotefair burden-sharingbetweengenerations.Instrumentssuch as revenue bonds, as well
as equity issues, can provide a good feedstock for emerging capital markets and an attraction to
institutionalinvestors. If specializedinstitutionsare createdto providecredit to municipalgovernment,
they should operate on solely financialcriteria. In manycases, local governmentsneed to improve their
creditworthiness.This may require more effectivetax mobilization,as well as changesin the sharing of
revenueand expenditureresponsibilitiesamonglevels of government.
Planning. The governmentoften must retain a role in planningof investmenteven where
executionof investmentand O&Mis providedby the privatemarket. Especiallywhere market conditions
remain very limited, it is necessaryto create measuresto ensure that activitiesof planning (as well as
regulation)are basedon incentivesfor efficiencyand responsivenessto demand.A demand-basedstrategy
for infrastructure development imposes significant information requirements: about the underlying
determinantsof demandfor specific services, users' willingnessto pay, and patterns of congestion,for
example.Performance indicators which reflect service quality and user satisfactionshould be used to
informdecisionson planningand policy-making,as well as the operationand regulationof infrastructure
activities. Since demand is ever-changing,it is necessaryfor infrastructuredecisionsto be based on a
dynamic process of assessing demand-and where competitive markets cannot provide this, other
approachessuch as regular demand surveysand channelsfor participationare needed.

1
Introduction
This paper concerns the choices policy-makers face in designing and implementing
institutionalarrangementsfor the provision of infrastructure-includingtransportation,electric power,
telecommunications,water supplyand sanitation,and irrigation.' "Institutionarrangements' are broadly
defined to encompassall the factors influencingthe environmentin which a project or organizational
entity operates. This environmentincludes the structure of ownership, horizontal relationshipsamong
entities(e.g., competition),hierarchicalrelationships(e.g., regulatorycontrols),financingresponsibilities,
and participationby various groups in decision-making.The classic notionof a "competitivemarket"
may be considered one type of institutional arrangement; there are also variations on this model,
involvingmanydifferent roles for the public and the private sector. The paper illustrateshow a variety
of institutionalarrangementsmay be applied to different infrastructureactivities, with the ultimate aim
of improvingthe performanceof infrastructureand its potentialcontributionto economicdevelopment.
Part One of the paperpresents somedecision-makingcriteria for assigningactivities"to the
market", and for determiningthe roles of governmentand the private sector. This section also outlines
a menu of institutional arrangements that represent different degrees of public and private sector
responsibilityfor the tasks involvedin providinginfrastructure.Part Two discussessome applicationsof
these arrangementsto infrastructuresectors in a number of countries, and provides a comparative
evaluationbased on these experiences.Part Three discussescross-cutting issues(regardingcompetition,
participation, risk-sharing, regulation, financing, and planning) which determine the incentivesand
conditionsfor good performanceunder any of these schemes.A Glossaryof the terms used in this paper
is provided in Appendix1.

These sectors embrace activities commonly known as 'public works" or 'economic infrastructure". Transport includes the
subsectors of roads and highways, railways, ports and waterways, urban transport, and road freight. The water supply and
sanitation sector includes both wastewater (sewerage) and solid waste management. This paper does not specifically address
social infrastructure, such as for health and education, nor certain urban structures such as public buildings, street lighting and
marketplaces, although some of the issues discussed here may apply to these activities as well.

3
I. A Framework for Analysis of Institutional Arrangements in Infrastructure
There has been much discussionin recent years about expandingthe involvementof the
private sector in infrastructure,and increasingexperimentationwith specific arrangementsfor doing so.
Often, however, these efforts are not based on an explicit, normative analysis of the rationale for
assigningspecificfunctionsto the public or private sector. One clear lessonfrom industrialorganization
analysisis that much canbe learned aboutthe actualstructure and performanceof an industry, and about
the potentialfor policiesto improvethese outcomes,by lookingcarefullyat the economiccharacteristics
of the goodsand servicesinvolved,the nature of production,and the structure of demand in the sector.
It is argued here that choosingamongvarious institutionalforms for infrastructureshould be based on
such an analyticalframeworkas describedbelow, which draws upon establishedconcepts in industrial
organizationas well as publicfinancetheory. The analysisfocuseson infrastructure,but couldbe applied
equallyto any goods or services.
This framework is based on the premise that the choice amonginstitutionalarrangements
should be based on the objectives of promoting efficiency, equity (fair access), and accountability
(responsivenessto users and other financiers) in the supply of infrastructureservices. There may, of
course, be other objectives, such as specific political concerns, which determinethe ultimate decisions
by policy makers. The discussion also starts from the premise that for goods and services generally,
competitivemarkets-mainly involvingprivate actors-are both the most efficientmode of supply and
most accountableto users' (consumers')needs.The task of the analysisis to identifywhere the conditions
of competitivemarkets apply, or can be approximated,in infiastructure activities; and where these
conditions do not apply, to identify the functions which government needs to undertake and can
implementeffectively.This analysis places the burden of proof on the policy maker to explain the
reasons for departing from competitivemarkets and the private sector for each activity involvedin
infrastructure.
To allocateresponsibilitiesbetweenthepublic and private sectors, one must first distinguish
amongthe functionsinvolvedin enablinginfrastructureto become available(sectoralplanningand policy
making, ownership,regulation, and financing)or in actuallysupplyingthe service (tasks of investment,
operation, and maintenance).Under this framework,the key questionis to determinehow extensively
government needs to perform the above functions; in other words, in which of these tasks is public
involvementwarranted, and through what kinds of policy instruments. International experiencewith
economicliberalizationduring the past decade, in infrastructureas well as in other sectors, has revealed
two importantconclusions:that the realm of economicactivity in which public interventionis required
is narrower than once believed; and that where it is warranted, the public role can often be exerted
through less distortingpolicy instrumentsthan those traditionallyused.
The intentionof this frameworkis to make the rationale for decisions as transparent as
possible.This analysisdoes not lead to one uniqueinstitutionalsolutionfor a giveninfrastructureactivity
in all contexts;rather, it suggestssome objectiveguidelinesfor designingpublic policiesto fit the specific
conditionsof infrastructureactivitiesin individualcountries.

4
A. Determining the "Marketability" of Infrastructure Activities
Since the efficiencycharacteristicsof a classicmarket are well understood,a good point of
departurefor decidingthe dispositionof any infrastructureactivityshouldbe the relevanceof the market
model-the burden of proof should be to identify where and why this cannot apply. Neoclassical
economics and public finance theory offers three grounds on which markets may not be considered
entirelyappropriateto supply infrastructure,and where "market failure" is expectedto occur:
The Nature of the Good2 Involved
A necessary condition for a market is that transactions can be made privately, between
individualbuyers and sellers. There are two criteria here:
(i)

Subtractability.A purelyprivate goodcan only be consumedby one person at a time.

Suchgoods(e.g., food and typicalconsumergoods)are said to be highlysubtractable
or rivalrousin consumption.At the oppositeextreme, purelypublic goods have low
subtractabilityor rivalry, becauseconsumptionby one individualdoes not lessen the
availabilityto others-such goods (e.g., clean air) are said to be jointly consumed.
A moreoperationaldefinitionof this criterionfor private transactionsis that the costs
of consumptionby a specific individual can be identified. For an infrastructure
facility with a limited physical capacity, such as a highway designed for a given
traffic volume, the marginal costs of each incrementaluser may be imperceptibly
small until the capacity limit is reached and congestionappears-but the fact that
congestion is possible indicates that the facility is, in principle, subtractable.
Similarly, scarce natural resources such as an aquifer should be characterizedas
subtractable,eventhoughthe marginalcostsof consumptionmay appearinsignificant
until the possibilityof depletionis actuallyfelt.

(ii)

Excludability.Individualconsumerscan also be excludedfrom transactionsinvolving

purelyprivategoods. Suchexclusionis not feasible(or excessivelycostly)in the case
of purely publicgoods.

In betweenthe two extremesof purelyprivate goods and purely publicgoods are toll goods
and commonpool goods.Toll goodsare characterizedby exclusionbut low subtractability(e.g., a piped
seweragesystem). Commonpool goodsare subtractablebut have a low feasibilityof exclusion(such as
many water and forest resources).
The two conceptsof subtractabilityand excludabilityare related, in that they both are based
on assessmentsof costs and benefits. Whether the costs of barring accessto additionalusers are worth
incurring depends on the costs they imposeon existing users. For example, once a roadway becomes
heavily congested,users may considerit worth the inconvenienceto install access gates and fees. The
public interest in policing the use of fisheries and forests grows as society becomes more aware of the
costs of resource depletion.There appears to be an increasingtendency for some goods which were
traditionallyseen as purely public (e.g., major highways)or common pool (natural resources) to be
viewed as increasinglyprivate or "toll-able"-both because the technologynow exists in some cases

The terms "good" and 'service" in this note are used interchangeably.

5
which makes it easier or cheaper to restrict access (e.g., electronic road pricing), and becausegreater
environmentalawareness is making society more cognizant of the costs of congestion and resource
depletion.
Appendix II presents a matrix which applies the above criteria to sixteen subsectors in
infrastructure.It becomes clear from this exercise that attemptingto determinethe 'marketability' of
infrastructureaccordingto these two criteria does not result in simple yes/no answers, but in degrees
(high/medium/low).Particularjudgementsin each case would depend in practice on the technologyin
use, which affects the marginal costs of consumption,the extent of congestibility,and the costs of
exclusion.Each subsector also consistsof various activities, each of which may have different degrees
of subtractabilityand excludability.The ratings for each activity and subsector from AppendixI are
summarizedon the followingmatrix (Table 1-1). The conclusioncan be drawn that, contrary to the
common assumption that most of infrastructure concerns largely "public goods", in fact the major
share of the services of these sectors can be characterized as closer to "private goods".
The Nature or Conditions of Production
(i)

Natural Monopoly.The efficiencyof the private market depends on the existenceof

effectivecompetition,which is precluded for goods which entail natural monopoly
in production.Naturalmonopolyoccursmainly where there are extensiveenomie
of scale (high fixed costs relative to variable costs, so that average costs decline
continuouslyover the relevant range of output).' Naturalmonopolyimpliesthat the
unit costs of supplying a given market will be minimized with a single
(profit-maximizing)supplier,rather than with multiplesuppliers. However, a single
or dominantsupplierwouldhave an incentiveto chargeprices in excess of marginal
costs, and so allocative efficiency would not be achieved. Moreover, lack of
5 Hence,
competitionmay blunt incentivesfor dynamic and productive efficiency.
where there is market power due to natural monopoly,the conditionsfor a purely
free private market to achieveefficiencydo not hold. Monopolypower also violates
the assumptionof "fairness" of transactionsand equity in relations betweenbuyer
and seller which underliesthe market model.

Where naturalmonopolyexists, it confersmarketpower and a natural barrier to entry. The

existenceof natural monopolyin a given activity depends on the cost structure, which varies with the
particulartechnology.Technologicaloptions may be availablewhich reduce economiesof scale and/or

' Appendix11(and the summaryTables 1-5) are intendedto be illustrative,not definitive,since there may be differentratings
in somecases basedon the particular technologyin use and other circumstances.
monopoly may also cntail economiesof scope (unit costs of producing different services in combination are lower than
of producing them separately), which continue to exist in some activities in power and telecommunications; and/or economies
of contiguity (the unit costs of producing services in markets which are located close to each other are less than if the market.
are servedseparately), which are a factor in solid waste collection, for example.
4Natural

5 This point has been debatedextensivelyin the literature (e.g.,

Scherer 1990).

TABLE 1-1. CHARACTERISTICSOF INFRASTRUCTURESERVICES:SUBTRACTABILITYAND EXCLUDABILITY

SUBTRACTABILITY
............

EXCLUDABILITY

COMMONPOOL GOODS

PUBLBCGOODS
LOW

Roadsand Highways: Tertiary roads - urban

Airpor/Airways: Air traffic control

Roadsand Highwevy: Signalingandtraffic

control
MEDIUM

Railways: Switching nd Signaling

Roadsand Highways: Primnry roads
(National, Trunk); Tertiary roads - rural
Solid Waste Manaement: Disposal:
Sanitary land fill
Urban Transport: Traffic signaling

Irrigation (SurfaceWatcr): Trunk System:Dam, main

canal; Distribution System:Secondaryand Tertiary canals
Roadsand Highways: SecondaryRoads
Solid WasteManagemeit: Collection
WastewaterManagement(Intermediate/LowCost
Sewerage):e.g. CondominialSewerage

Irrigation (SurfaceWater) Terminal System: Gravity

Water Supply (piped): CommonTerminal Equipment(i.e. handpump)

HIGH

WastewaterManagement(Conventional
Sewerage):ConventionalStreetSewer;
PumpingStation; TreatmentPlant

Airports/Airways: Runways,gates
Ports and Waterway: Piers, harbor
Railways: Railbed
Solid WasteManagement:Disposal: Incineration
Telecommunications(Basic Network): Networks
Transmission:Local, Long Distance;Switching; Common
Terminal Equipment
Urban Transpor Metro/rapid tranit: Tracks andRails

Airports/Airways: Airplane Services; GroundServices

Electric Power: Generation:thermal, hydro, nuclear;Transmission;
Distribution
Irrigation (Groundwater):Tubcwells
Irrigation (SurfaceWater): Terminal System: Sprinkler
Ports and Waterways: ships; Port equipment(Loading/Unloading)
Itailways: Rail cara; Freight Loading/Unloading
Road Freight Transort: Trucking Services
Solid Wastemanagement:Transfer; RcaourceRecovery(recyclng)
Telecommunication(Basic Network): Individual Terminal Equipmet
Teleconmuunications
(network Extensions):Value-AddedServices;Cellular,
Paging, Microwave Relay; Private or SpecializedNetworks
Urban Transo r: Van, bus; Taxi; Metro: Railcara
Wastewatermanagement(Intermediate/LowCost Seweraec):BasicSanitation
(pit latrine)
Water Supply (Nonpiped):Vendor tanks; Boreholes
Water Supply (piped): Trunk and Distribution System;Individual Terminal
Equipment(i.e. home faucet)

TOLL GOODS

PRIVATE GOODS

Definitiona:
Subtracrabliry (or rivalry) refersto the impact that consumptionby incrementalusershason consumptionopportuniticsof all users.Low subtraciabilgy (nonrivalry) - consumptionby one user doesnot inpede
availbility to other usera(suchgoodsaei 'jointly consumed').High subrraciabliry = consumptionby one user imposesidentifiable, calculablecoats on other users(suchgoods ar 'privately consumed').
EclsudabliWtyrefersto the feasibility of controlling accessto a good. High erxludabiliry meansit is rlatively easy(t.e. non-costly)to prvent userafrom consumingit.

os

7
reduce the natureof capital investmentneededto enter the market. Recentadvancesin economictheory'
have contributedto the understandingthat the main deterrentto competitionis the elementof the fixed
productioncosts which wouldbe lost in the eventof an unsuccessfulattemptto enter the market-in other
words, the magnitudeof sunk costs in the event of exit. Capital costs are "sunk" to the extent that they
cannot be recovered for other uses, which is generallythe case for specializedequipment(i.e., firm,
rather than industry, specific)and fixed (location-specific)installations,such as roads and sewer pipes.
Whereproductionof a good requiresno sunkcosts, it is said to be perfectlycontestabie,and the potential
threat of entry is consideredto provideapproximatelythe samemarket disciplineto an existingincumbent
as actual competition.There may be still be practical barriers to entry imposed by policies or other
factors such as shortagesof financing;these barriers may be addressedseparatelyas long as an activity
is contestablein principle.
Thematricesin AppendixII characterizethe activitiescomprisingeach infrastructureservice
subsector by the degree of economiesof scale (for the types of technologymost generally in use) and
"sunkness" of costs of capital involved. These characteristicsare summarizedon Tables 1-2 and 1-3
below.
(ii)

Coordination.A concept which is relatedto, but distinct from, natural monopolyis

coordination.Becauseof the interlockingnetworksinvolvedin muchof infrastructure
and the fact that flows along these networks (whether of electric power, voice
signals, water, or vehicle traffic) need to follow some explicit rules, formal
coordinationis required to some degree-at least in the planning of investment,
technicaloperation, and setting of minimumstandardsof equipmentand operation.
The extentof coordinationneededby the various infrastructureservice sectors is also
indicatedin AppendixII, and summarizedin the followingTable 1-4.

Externalities and Social Objectives

Externalitiesoccur where the benefits (or costs) of producingor consuminga good affect
persons other than the individualsinvolvedin the transaction.When the externalitiesare positive, the
benefitsto societyare greater than thoseperceivedby the individuals,and the resourcesallocatedto such
goodswill be less than socially optimal.Whenthe externalitiesare negative,the marginalcosts facedby
individualsunderstatethe true cost to society, and such goods will be overproduced.In infrastructure,
negativeexternalitiesincludeeffectson air, water, and land pollutionfrom motorvehicletraffic, electric
power generation, and irrigation, while positive externalitiesinclude the public health benefits gained
from water and sanitationinfrastructure.Many infrastructureactivities also pose external economiesof
consumption(networkexternalities\,wherebyall users (e.g., of voice telecommunications)benefit when
a new subscriber gains accessto a networkbecause it increasesthe range of service they can all enjoy
-namely, the abilityto communicatewith more people.
In a similar vein, certain social or political objectivesvalued by the community,such as
universal access to a minimumlevel of service (particularlyfor water and personal mobility) may be
consideredto have a socialbenefitbeyondthat accruingto individualusers. Suchso-called "merit goods"
have positivesocial externalitiesof consumption.

6 Baumol, William J., Panzar, John C., and Robert D. WiJlig, Contestable Markets and the Theory of Industry Structure. San

Diego: Harcourt Brace Jovanovich, 1988.

8
The following Table 1-5 summarizes these types of considerations for the various
infrastructureservices, as identifiedin AppendixII.
The three criteria discussedabove (privatevs. public nature of goods, natural monopoly,
and externalities)are normallyconsideredto cover the conditionsfor "market failure", i.e., the inability
of the market model to achieveefficiencybecauseof characteristicsof the goods and their production.
A fourth set of considerationsconcerns the nature of demand for the goods. These features suggest
additional requirements for consumers to acquire satisfaction from any given supply
arrangement-requirementswhich may have implicationsfor regulatoryor price policy, as discussedin
SectionB and more fully in Part III.
Characteristicsof Demand and ServiceUse
(i)

Existence of substitutes. When there are acceptable and affordable substitutes

availablefor the services providedby a main supplier, the latter will be unable to
exercise the same market power as a true monopolist.Which alternativesare the
relevant substitutesis a judgement made by the consumers, and these alternative
servicesneed not be identicalto those of the dominantsupplier. For example,when
the publicpower sourcebecomestoo unreliable,firms turn to privategenerationeven
when it is at a higher cost to them. Householdssimilarly resort to water vendors
when the public water utility fails to serve them adequately. Specialized
communicationsnetworks for high-volume business users are sought in some
countriesto bypassthe congestedpublic telephonenetwork. Much shipping can be
loadedand unloadedby lighterswithoutenteringport facilities, for whichthe lighters
serve as a partial substitute. In many countries, freight traffic by road has become
the preferredsubstitutefor rail transit. Since muchof infrastructureis fixed in place,
consumers-especially the poor-are often physically restricted in their access to
substitutes(e.g., the roads have to exist in the same regions served by rail). Thus
substitutabilityis specificto locationand to priceand incomevariables.The issue
becomes particularly important for determining the welfare costs of monopoly
suppliersof many infrastructureservices.

(ii)

The price elasticityof demand.The classicmarket modelassumesthat the consumer

is able to adjust demandto price. However, for certain goods in certain minimum
quantities(drinkingwater beingthe mostobviouscase), demandis virtuallyinelastic,
meaning that that quantity would be consumedat any price. Beyond this extreme
case, the price elasticityof demandfor infrastructureservicesvaries greatly among
different groupsof consumers.As a generalization,minimumlevels of consumption
of urban public transport, electricity, and even basic voice telecommunications
demonstrate relatively inelastic demand as development increases-they become
"necessities" to gain access to jobs, health care, etc. Wherever demand is
price-inelastic,marketpower by suppliersposes a greater risk to welfare.

TABLE 1-2. PRODUCTION CHARACTERISTICS

OF INFRASTRUCTURE ACTIVITIES: ECONOMIES OF SCALE

ACTIVITY

ECONOMIES OF SCALE

SECTOR
Airports/Airways

LOW
Airplane Services; Ground services

Electric Power

MEDIUM

HIGH

Runways, gates; Air traffic control

Generation, Distribution

Tranamission

Irrigation (Groundwater)

Deep Tubewell; Shallow Tubewell

Irrigation (Surface Water)

Terminal System (On-farm): Gravity,

Sprinkler

Trunk System: Dam, main canal;

Distribution Sytem: Secondary and
tertiary canals

Ports andWaterways

Ships; Port equipment (loading/unloading)

Piers, harbor

Railways

Switching and Signaling; Rail cars; Freight

Loading/Unloading

Railbed

Road Freight Transport

Trucking Services

Roads and Highways

Primary Roads (National, Trunk); Secondary

Roads; Tertiary roads - rural, urban;
Signaling and traffic control

Solid Waste Management

Collection; Resource Recovery (recycling)

Transfer; Disposal: Sanitary land fill,

Incineration

Telecommunications (Basic Network)

Network: Transmission - Long Distance;

Switching; Terminal Equipment

Network: Transmission - Local

Telecommunications (Network Extensions)

Value-Added Services; Cellular, Paging,

Microwave Relay, Private or Specialized
Networks

Urban Transport

Van, bus (urban, inter-urban); Taxi;

Metro/rapid transit - Rail cars; Traffic
signaling

Wastewater Management
(Conventional Sewerage)

Metro/rapid transit - Tracks/Rails

Conventional Street Sewer, Pumping

Station; Treatment Plant

WastewaterManagement (Intermediate/Low
Cost Sewerage)

Condominial Sewerage; Localized treatment;

Other intermediate sewerage;basic sanitation(pit latrine)

Water Supply (Nonpiped)

Vendor Tanks; Borehole

Water Supply (Piped)

Terminal Equipment: Common (i.e.

handpump), Individual (i.e. home faucet)

Trunk System (intakepumping sation);

Distribution System

'

TABLE 1-3. PRODUCTION CHARACTERISTICS OF INFRASTRUCTURE ACTIVITIES: SUNKNESS OF COSTS

ACTIVITY

SUNKNESS OF COSTS

SECTOR
Airport/Airways

LOW

MEDIUM

Airplane Services; Air trffic control; Gmund

Electric Power

HIGH
Rumvgys, ge

Generation, Distribution, Transmission

Irigation (Groundwater)

Shallow Tubewell

Deep Tubewell

Irrigation (Surface Water)

Terminal System (On-farm): Gravity,

Sprinkler

Trunk Sytem: Dam, main canal; Dirinbution

System: Secondary andtertiary cnals

Ports and Waterways

Ships

Port Equipment (oading/ unloading)

Picrs, harbor

Railways

Switching and Signaling; Rail cars

Freight (loading/uhloading)

Railbed

Road Freight Transport

Trucking Services

Roads and Highways

Signaling and traffic control

Solid Waste Managemeit

Colection; Tranfer, Resource Recovery

Disposal: Sanitary ln
Incinerationl

Telecommunications (Basic Network)

Network: Tranmission - Long distance;

Switching; Terminal Equipment: Individual

Terminal Equipment: Common

Telecomrmunications(Network Extensions)

Value-Added Services; Cellular, Paging,

Microwave Relay; Private or Specialized
Networks

Urban Transport

Van, bus (urban, inter-urban); Taxi; Traffic

Sigaling; Metro/rapid trnit: Rail cars

Primary Roads (National, Taunk); Secondary

Roads; Tertiary Roads: rural, urban
fll,
Network: Transmission - Local

Metro/rapid transit: Trck/Rails

Wastewater Management
(Conventional Sewermge)

Conventioral Street Sewer, Pumping Station;

Treatnent Plant

Wastewater Management (Intermediate/Low

Cost Sewerage)

Basic Sanitation - (pit latrine)

Water Supply (Nonpiped)

Vendor Tanks

Borehole

Water Supply (Piped)

Terminal Equipment: Common (i.e.

handpump), Individual (i.e. home faucet)

Trunk System (tntake pumping station);

Distribution System

Note:

High sunk costs imply low contestability, and vice versa.

Condominial Sewerage; Localized

treatment; Other intermediate sewerage

TABLE 1-4. PRODUCTION CHARACTERISTICS OF INFASTRUCTURE

ACTIVITY

ACTIVITIES: COORDINATION REQUIREMENTS

COORDINATION REQUIREMENTS

SECTOR

LOW

Airport/Airways

MEDIUM
Ground Services

Electric Power

HIGH
Airplane ervices; Runway, gates; Air traffic
control
Generation; Trmniuion;

Distibutios

Irnigation (Groundwater)

Shallow Tubewell

Inigation (Surface Water)

Terninal System (On-farm); Grvity,

Sprinkler

Trunk System: Dam, main canal; Distribution

System: Secondary and tertiary canals

Potts and Waterways

Ships, Post equipmedt (loading/unloading)

Pien, harbor

Railways

Deep Tubewell

Freight Loading/Unloading

Road Freight Transport

Trucking Services

Roads and Highways

Primary Roads (National, Trunk); Secondary

Roads; Tertiary Roads - rural

Solid Wase Management

Resource Recovery (recycling)

Telecommunications (Basic Network)

Railbed; Switching and Signling; Railcars

Tertiary Roads - urban; Signaling and traffic

control
Collection

Transfer, Disposal: Sanitary land fill,

Incineration

Terminal Equipment

Network: Tnsnmission - Local, Long

Distance; Switching

Telecommunications (Network Extensions)

Value-Added Services; Cellular, Paging,

Microwave Relay; Private or Specialized
Networks

Urban Transport

Van, Bus (urban, inter-urban); Taxi

Watewater Managemnent
(Conventional Sewerage)

Pumping Station; Treatment Plant

Wastewater Management (Intermediate/Low

Cost Sewerage)

Basic sanitation (pit latrine)

Water Supply (Nonpiped)

Vendor Tanks; Borehole

Water Supply (Piped)

Terminal Equipment: Common (i.e.

handpump), Individual (i.e. home faucet)

Metro/rapid transit: Tracks/Rails, Railcars;

Traffic signaling
Conventional Street Sewer

Condominial Sewerage; Localized treatment;

Other intermediate sewerage

Trunk System (intake pumping station);

Distribution System

12
Table 1-5. EXTERNALMES AND SOCIAL OBJECTIVES PERTAINING TO INFRASTRUCTURE
Socio-political Objectives
Externalities
Sector/Subsector
Transportation
Railways

Urban Transport
Bus, taxi

Rapid transit

Network effects'

Access to service for

remote areas

Air pollution
Urban congestion
Public safety
Noise

Affordable access to means of

personal mobility

Urban land use

Road freight

Air pollution
Noise

Common carrier
Access for remote areas

Roads and Highways

Affects settlement/land use

patterns, drainage, erosion
public safety, dust pollution

National integration, access to

remote areas

Airports

Noise, public safety

National defense

Ports/waterways

Water pollution

National defense

Electric Power

Air pollution (thermal)

Radiation (nuclear)
Resettlement (hydro)

Affordable access to minimum

service

Telecommunications

Network effects

Affordable access to minimum

service
National integration

Water Supply

Intersectoral allocation of
water resource;
Public health

Affordable access to
minimum service

Sewerage

Land, water pollution

Public health

Affordable access to
minimum service

Solid waste management

Land/air/water pollution,
Public health
Drainage

Affordable access to
minimum service

Irrigation

Intersectoral water allocation;

waterlogging, salinization, erosion
of land; depletion, pollution of aquifer;
introduction of diseases

of consumption(decisionby one user to link to the networkyieldsbenefitsto other users,

'Networkeffects-Refersto positiveexternalities
throughthe telephonesystems).As anotherexample,for a railwaybetweentownsA
e.g. by increasingtherangeof possiblecommunications
andB, benefitsto useroat anypointalongthe line increasewhena linkto townC is added.

13
(iii) Consumers' access to information.The premise of "consumer sovereignty" in the
market is based on consumershavingaccessto informationabout the characteristics
of products and services.In some areas of infrastructure,the qualitiesof service are
not easily assessedby the consumer(e.g. purity of the water supply), which makes
it more difficultto "shop around" and thereby challengethe dominanceof particular
suppliers. Anotherexampleof asymmetriesin the informationavailableto producers
and consumersis providedby the urban transport sector. Individualscannot assess
the safety of the buses or taxis in which they ride, which justifies public safety
regulation.In addition,inefficiencies(in the form of suboptimalprice and frequency
of service)arise in urban bus markets wherethere is fullyunregulatedentry, because
of imperfections in consumers' knowledge about the actual service alternatives
availableat any one time.' Indicatorsof servicequalitywhich make the performance
of services more transparent increase the "marketability" of infrastructure.
Similarly,the prospects for providingservices efficientlythrough a market can be
enhanced when informationis made publicly available regarding the impacts of
consumptionon nonusers(i.e., where there are externalities,such as air pollution
causedby motor vehicle traffic).
(iv) Temporalpatternsof demand. For certaininfrastructureservices(especiallyelectric
power and voice telecommunications,and to a lesser extent, water supplyand urban
transport),demandis not randomlydistributedover time but showsdistinctpeak and
off-peak periods. The outputsof some of these servicesalso cannotbe readilystored,
and the capacityof the system needsto be relatedto peakdemandrather than average
demand. Where facilities involve large, "lumpy" capital investments,the physical
capacity cannot readily be expandedor contracted and the supplier must bear high
costs of excess capacity in slack periods, as is the case for example with power
transmission.However, where incrementalfacilitiescan be brought into service to
meet peak demand (such as from standby generators which are connectedto the
transmissiongrid), this may be more efficientin some cases than addingto the main
capacity.Likewise,at some times, constructionof extra capacityto avoidcongestion
at the peak periods may be avoided by charging higher prices for peak than for
off-peak consumption.Removal of regulations which impede such approachesto
dealingwith peaks in demand can reduce the need for very large investments.
(v)

Extent of diversity of user needs. A traditional conception about much of

infrastructureis that it produces fairly homogeneousproducts (e.g., generic power,
telephone,freighttransportservice)whichlendthemselvesto standardizedproduction
processesby centralpublic suppliers. Yet it is becomingrealized from research that
evenhouseholddemandfor drinkingwater is not uniform but varies with perceived

7 The peculiar features of urban transport markets in this respect are described in Dagerman (1992).
8 A relatedpoint is that a condition for production by profit-seeking agents is that the requirements of the product'service can

be fully defined in advance, and that the product delivered can be evaluated unambiguously. These conditions are not met if
information about service performance and quality is not readily available to users (John D. Donahue, The Prvatizadon
Decision: Public Ends, Private Means. New York: Basic Books, Inc., 1989)

14
qualitydifferentialsand reliability(e.g., Humplickand Madanat, 1992). Moreover,
user needs for telecommunicationsand transportare becomingever more diversified
due to rapid changes in technology and production processes. The increasing
differentiationof user demand impliesfor some servicesthat the structureof supply
has to become more diverse as well to respond adequately. Thus, this last
characteristicof user demandprovides an argumentfor multiplicityof suppliers.
To concludethis sectionon the "marketability"characteristicsof infrastructuresubsectors,
it can be said that the provisionof network (pipeline-type) facilities, especiallythe primary or "trunk"
level (e.g., basic transmission hardware for power, telecommunications,piped water supply and
sewerage,trunk highways,port berth structures,and large scale irrigationnetworks)exhibits,to varying
degrees, the characteristicsof public goods, natural monopoly,and high sunk costs. The operationof
these networks,on the other hand (tasks involvedin generatingservices from the facilities), oftendoes
not entaillarge sunkcosts for equipment,and is thus contestable.In additionto these features, it has been
arguedhere that the nature of externalitiesand social values associatedwith the services, as well as the
nature of the market for the servicesfrom the users' perspective,must be understoodin order to identify
the appropriate roles that the public and private sectors should play concerning both facilities and
operationalfunctions.
In manycases, technologicalchangehas transformedthe economicconditionsof production
and the nature of demand for the servicesaccordingto the abovecriteria. Technologyhas also thereby
introducednew optionsfor supply (see Box 1).
The key point of this sectionis that to properlyjudge the "marketability"of infrastructure,
it is necessaryto considerthe economicand technologicalcharacteristicsat the level of specificactivities
within subsectors. Looked at in this way, as illustrated by the tables above, it becomes clear that
infrastructureis highlyheterogeneousnot only acrosssubsectors,but also withinsubsectors.Taking
the airlines subsector as one example, air traffic control is a pure public good; runways and gates are
essentiallytoll goods; and airline and ground services are private goods. Among the infrastructure
activities charted on these matrices, the majority appear as private goods rather than as public goods.
Fewer of the activities involve large economiesof scale than is customarily imagined; however, a
significantshare of the activities involvehigh sunk costs and require technical coordinationto operate
efficiently-featuresthat call for some regulationwhen private entry is allowed.
B. Determiningthe Public and Private Roles in Infrastructure
Oncethe basic "marketability"of each activityhas been identifiedin this way, it is possible
to assign responsibilitiesfor specific functionsbetween the public and private sector. For any service
involvedin infrastructure-or for that matter, any other sector-if all the conditionsfor a competitive
market as just discussedare met, then it is unquestionablypreferable from an efficiencystandpointto
provideit throughthe market.However,becauseof the characteristicsnotedabove, most of infrastructure
cannot be served by private markets. Governmentinterventionis justified in, and should be limited to,
cases where the potentialcosts of market failure are greater than those of governmentfailure.

15

Box 1 The Changing Technologyof Infrastructure

Althoughnot all the technologicalchanges relevant to this discussionhave appeared during the pas.
decade, their implicationshave begun to be widely exploitedonly ia recent years as they provided support for the
concurrentworldtrend towards economicliberalization.The followingbriefly illustratesthe significanceof technological:.
alternativesfor creatingnew supply options and for changesin demnandfor infrastructure:
reducing conditions for natural monopoly. In telecommunications,technologicalchange haxsreduced economiesof scale in long-distancetransmission,underminingnaturalmonopolyin this area. In local exchangui
service, new transmissiontechnologiessuch as cable-basedtelephoneaccess, cellularradio, and directmicrowavecreatc
some de facto competitionwithin the local exchangemarket. Digitalizationof switchinghas simplifiedmaintenance
functions,thus reducingeconomicsof scale in overheadactivitiessuch as O&M. In electricpower, technologiessuch as
gas turbine generatorshave also reducedeconomiesof scale in generation.
permitting low-costsupply options. Intermediatesanitationtechnologieshavelowerconstruction
costs than conventionalsewerage, makingthem affordablefor low-incomecommunities.Changesin designparameters
for conventionalsewerage, based partly on technologicaladvances,have also permitted lower-cost alternativessuch as
condominialsewerageto be used where communitiesorganize and participatein planningand implementation.Among
alternativesto traditionallarge surface schemesin irrigation,options such as drip, bubble, and sprier systems and..
low-levelcanalswith low-lift pumps are highly responsiveto demandfor water. They promoteconservationand foster::
privatemanufactureand ownershipof the equipmentinvolved.

inereasing range and quality of service. Value-added services in telecommunications(e.g.

facsimile),whichare the most dynamicsourceof demandinthis sector, combinetransmissiontechnologieswith computer
processing.The containerrevolutionin transportationpermitsrapid and cost-effectivetransfer of freight across multiplc
transport modes. Combinedwith electroniccommunicationssystems, intermodaltransporthas greatly reduiicdtransportt..:
costs and improvedthe qualityand speedof trade logistics.
facilitatingthe "unbundling" of assets and operations. Technologiesof "nondestruetivetestin.'
aLnd
remote monitoring(c.g., use of cameraswithin pipelines)perrnitthe conditionof fixed infrastructurefacilitiestobe'..
assessedand problemsdiagnosedwithoutcostlyand time-consumingexcavationor dismantling.Thiscapacityalso implies
hiat where the ownership of fixed assets is vested in a different entit than operation (e.g. in a regional wate.
supply/sanitationsystemor toll road), the owneror regulatoryauthoritycanindependentlymonitorthe conditionof these
assets.

. ' .:: ' --. .,. , . ..

..

. . . ..

*
e~xpanding
options for demand management. In the roadssubsector,clectronicroadpricingcani'...
permit road user charges to be differentiatedso as to reflect the actual impact of diffcrent vehicle loads on road:
deterioration,to managecongestion,and to internalizethe social costs of pollution. Automaticvehicle identification-s
alreadyusedan manytoll roads in the UnitedStates, Norway,and elsewherein WcsternEuropeand ixplannedin Mexico
and Singapore.

More analysishas been devotedto the sourcesof market failurethan to those of government
The latter may include factors such as unresolvableconflicts among policy objectives; the
inabilityof policy-makersto interpretthe "publicinterest" accurately;and the interplayamongthe public
interest, the concernsof particularconstituencies,and the private interestsof officials.It is arguedhere
that to minimizethe risks of governmentfailure, it is necessaryto be very clear about the argumentfor
assigningany specific task to government;moreover, once it is determinedthat governmentshould have
failure.9

9The literatureon publicchoiceand principal-agenttheories,for example,and other institutionalanalysishave examinedissues

of governmentfailure.

16
responsibilityfor a particular activity, it is preferableto choose the least interventionistpolicy response.
In determininga strategyfor transitionto this target, it will of course be necessaryto take accountof the
constraintsof social, political, and institutionalcircumstances(e.g., human capital) in each situation.
In many cases, argumentswhich have traditionallybeen presented for public ownership,
financing, and/or delivery of infrastructureservices can be addressed by a more restrained policy
response to achievethe same efficiencyand equity objectives.This approach can be illustrated very
brieflybelow,by relatingthe specificreasonsfor governmentinterventionto a menuof "minimal"policy
actions:
(i)

For activitiesinvolvingpublic or quasi-ublic goods, natural monopoly, or capital

with high sunk costs-typically, the provision of network, trunk-type facilities as
discussedearlier-there is a case for publicplanning and policy making, as well as
public financingand ownership. An alternativemay also be private sector financing
and ownershipunder public regulation.

(ii)

Sincethe activitiesinvolvedin operatingsuch facilitiesare normally contestable in

nature, there is no reason a priori why the public sector should operate such
facilities. Operationand maintenancecould be ailocatedon the basis of competitive
biddingfor the right to the monopoly. The government'sresponsibilitiesin this case
are to issuethe exclusivecontractand monitorperformanceunder its terms; to ensure
that other providers of services using the facilities face fair conditionsof access
(includingprice); and to protect users from other possible abuses of the monopoly.

(iii) For other activities that do not involve high sunk capital costs (e.g., road freight
transport), there is generallyno economicjustificationfor policiesthat impedeentry
by the private sector.' 0 ' The governmentshouldmainly be responsiblefor ensuring
fair competitionin such activities.
(iv)

Externalities, such as environmentalimpactsof infrastructure,can be addressedby

regulations (e.g., on zoning, technicalstandards)or fiscal transfers (taxes, fees, or
subsidiesto influenceprivate investmentor operation).'2

(v)

Distributional objectives can be met by regulation (e.g., universal service

requirements),investmentplanning(e.g., to ensure a certain regionaldistributionof

'As indicatedcarlier, a qualificationis neededfor urbanbus transport,where someregulation(e.g., in the form of competitive

licensingof routes)is justifiablebecauseof structuralfeatures of the urban bus market.

11Even where a particularactivity does not entail sunk costs, if there are economiesof integrationor scope with the other
setiviticaundertakenby a dominantprovider(e.g., the railwayor water utility),this fact may impedethe ability of new entrants
to oompeteon an equal cost basis. In such cases, there may be need for competitiverestructuringof the monopolist-meaning
horizontaland vertical separationof the contestableactivitiesfrom those involvinglarge sunk costs.
12 Th larger the cxternalities (positiveor negative),the morejustificationfor public involvement.As put by Donahue(op
cit.), a poorly-performingcontractorshould facepenaltiescommensuratewiththe potentiallossesit inflictson the public. When
thesnloues are very great (e.g. of a nuclearpower plant disaster,or breakdownin air traffic control), there is justificationfor
more publicinvolvementin the activity.

17
essential minimumservices), and/or public financing of nonremunerativeservices
which are deemed of social importance.
(vi)

Significantrequirementsof coordination among facilities and services may justify

regulationof investmentor operatingstandards.

(vii) Characteristicsof market demand, such as low availability of substitutes or low

price elasticity may arguefor regulationof tariffs to protect consumersif the supply
is a natural monopoly.Assymetries in information may justify regulationsuch as
requirementsregardingdisclosureof informationaboutservice qualityand alternative
sources of supply. Where temporal patterns of demand exist, there may be a case
for some public planning of investment,deregulationto promote efficient use of
reserve capacity, and pricing policy to shift demand to off-peak periods. A high
diversity of services demanded would suggestpoliciesto liberalizeproductionand
promote competition. It is worth noting that none of these considerationsof
externalitiesor characteristicsof demandprovidesapriorijustiricationfor public
ownership,nor even for the public sector to engage in all the tasks of planning,
regulating,financing,and deliveringthe services.
The analysisto this point has referred to assigningfunctionsto the public and private sector
to meet the objectivesof efficiencyand equity. A third objective mentionedearlier is accountability,
definedas the abilityof serviceproviders to serve the interestsof users and other financiers.Competitive
markets meet this criterion of accountabilityin that the profit-oriented suppliers have an incentiveto
satisfy demandsof their customers(and it is customerswho provide the revenues). Where competitive
markets cannot be used to provide infrastructureservices, other arrangementshave to be found to
3 Such measures may include decentralization of activities undertaken by
promote accountability."
government administration, if such decentralization promotes greater responsiveness to users;
participation by users in theplanning,financing,and deliveryof infrastructureservices,whetherthrough
"self-help" or community forms of organization, or through other channels for interest groups to
influencethe decisionsof service providersand regulators.It is also importantthat indicatorsof service
quality and other dimensionsof user satisfactionbe incorporated into the processes of planning and
regulation.

C. Characteristicsof AlternativeInstitutionalArrangements
The analysisabovehelps to narrow the range of publicpolicy options to be consideredfor
particular activities, but does not lead to one unique institutionalsolution for a given infrastructure
activityin all contexts.The range of institutionalalternativesis in fact broaderthan is often recognized.
The options fall along a continuumbetween the extremes of completelypublic sector and completely
private sector responsibility,as shown in Table 1-6. From among these options, specific forms can be
identifiedas appropriatefor particularactivities.

13 Wheneverservicesare not financedby user charges, there can be a conflictbetweenthe suppliers' incentivesto satisfythe
users, and the need to satisfy those who provide the revenues(e.g., often the governmentbudget). Thus, in additionto the
measures listed here to increase accountability,it is importantthat infrastructureservices be subject to user charges to the
greatestpossibleextent (see 'Pricing and Financing"in Section111.)

18
Table 1-6. Types of InstitutionalForms by Degree of Publicand Private
SectorResponsibility'
Rangeof Responsibility
.

.v

PUBLICSECTOR
xxxxxxxxxx
xxxxxxxxx
cxXX XxXXService
X
xxxxxx

xxxxxX
O

-v

Menu of Institutional Oplions

xxxx
XX
x

GovernmentDepartment
Parastatal
Contracting
ManagementContracting
Lcasing
Concessions(including BOTIBOOetc.)
Cooperativelcommunalarrangements
2
Privatc cntreprencurship

PRIVATE SECTOR
1. SecAnncx111
andCloas%ary
for &ila dcfinaiWons,
anddu=rip1ions.
ithscai.egoriCs
arc nolmuntiialy
cxc]uswvc,
c.g.g'c1rncjtdepantnxriUs
andpa2sUals
nmav
cmploy
thr vaiousfonns
ofe
etianUg
aLX wpcr4fics andpdvac cnuvcprcncuw5
my wwi un&kr
cWuD-nlia
schmc. fr, publicWC*uI
cntiics.
2. Withmajoritylo rsra privateshareboldina.

Table 1-7 providesa moredetailedtaxonomyof the institutionalalternativeswhich indicates

howthey comparein their assignmentof functions,risks, and compensationto public and private parties.
Table 1-8 illustrateshowthe variousinstitutionaloptionsmay applyto particularinfrastructureactivities,
after taking into account the extent to which each activity lends itself to market conditions and the
requirementsfor governmentinvolvement.If the aim is to choose alternativeswhich promote efficiency
and accountability,then those which involve greater degrees of private sector responsibility and
involvementby beneficiariesi.e., those towards the right-hand side of Tables 1-7 and 1-8 would be
preferred whereverfeasible.
The next section describesthese arrangementsas they are currently in practice in various
infrastructuresectors, with an emphasison developingcountries.AppendixIII includesa broader(though
not exhaustive)listing, by infrastructuresector and country, of where the main alternativeforms are
planned or under implementationat the present time.

TABLE 1-7. COMPARISON OF INSTITUTIONAL FORMS BY ASSIGNMENT OF

FUNCTIONAL RESPONSIBILITIES
FORMS

'Traditional"

Department

FUNCTIONS

Cormneriafized

Management

Leasing

Concession

Private

'Self-Help'

Contracqing

(Affermage)

(BOT)

Entrepreneur

Cooperative/

lUI
~~~~~~~~~~~~~~~~~~~~~~
Co

State

Sectoral Investrent
Planning and PolicyMaking
Regulation

Internal
to
Govermnient

By
Parent
Agency

Capital Financing
(Fixed Asset)

Governnent

Large
Recourse to
Subsidies and

Current Financing
(Working Capital)

Govermnent

Governmentbacked
ing

__ _ _ __ _

or

or wn

Parastatal
(linited
aubsidies;
Market-based
financing)
I_I_
_ _ _ _ __ _ _

Execution of Investment
By Parastatal

Governnent
Government

or

State

State (Majority)
State

Operation a)
Maintenance

Contracting

Parastatal/Public Utility

Ownership of Assets

l ___________________

Service

| Corporatized

Govermment

Mixed

Private
(Majority)

Private or in
common

Public
Authority
negotiated
w/Contractor

None or Public
Authority

None or
Public
Authority

Parent
Separate

agency
Public

Authority

Public

Public

Public

Private
Contractor

Private

Private

Public

Public

Private
Contractor

Private
Contractor

Private

Private

__ _ _ __ _

_ _ __ _ __borrow

Private
Contractor
for Specific
Services

Private
Contractor

Public Partner
______

Private

Private

---------------

Private Contractor

Other Characteristics:

Government

Government

Parastatal

Public
Partner

Governnent

Government

Parastatdl

Public

Managerial Authority

Commercial Risk

I
I

Private
Contractor

Private
Contractor

Private
Contractor

Private

Private

Mainly Public

Private

Private

Private

Private

Contractor

Contractor
Privately
Determined

Privately
Determined

Partner
Basis of Private Party
Compensation
1-------------------

Duration
l

_ _ _ _ _ _ _ _ _

N/A
-------------------

No limnit

N/A

N/A

Based on
services
rendered

------------------------------

No limnit

No limit

Less than 5

_ _ _ _ _ _ _

years

Based on
services and

Based on results, net of

contractor payment for use of

results

existingassets

---------------------

Jyears

About 3-5

------------------------

------------

10-30 years

5-10 years
_

_ _

No linmit

No limit
_

_ _

_ _

_ __yee

l__

TABLE 14.

j Government

FORM

INSTITUTIONAL FORMS APPLICABLE TO

INFRASTRUCTURE ACTIVITIES

Parastata/

Department

Publk

Nationaior

Utility

j Concession

Mmnageint
Sersike
Contract

Contract
_

Leasing

| BOT|

_o_(Affennaej

I.

Sub-National)
S
SECrOR

i*)

POWER:

Regulted/

Entrepreneur
;Conerad've

Unregulated

Entpreper

Coupentive
Co unal

Overanl

Generation

Small Scale
Commrunal

TranamissionSysem

Enclave(

System

sactord
Distribution

Systemr

Local transmission, swit.ningsysem

TELECOMS:

Long Distance

Transmission

Terminal

Comnunal

Equipment

Systems

service
Extnson

and
TRANSPORT:
polcy

Passenger andFreight Rail

Enclave'

RAilways

Urban
Transport

making

Urban
Bus

Commuter Rapid Transit

Freight

..... :.,..

Taxi, Van
Inter-Urban Bus

.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....

FRreight

Roads nO
Roads and

Prinmary, Secondary, Urban Roads

|Toll

Road

Enclavof

Highways

Rural and

I Local Roads

Airport/
Airlines

General-Use Airports

Enclave and Airlines

Enclave! and Shipping

Ports and
Waterways

General Use Ports

WATER &
SANITATION
Water Supply

Pipe Water Trunk & Distribution

Enclave'
Shallow Well;
Vendor

Sewerage

ConventionalSewerage and Treatmnent

Solid Wase
Management

IntermediateSewerage

Collection/Transfer/Dispoaal

IRRIGATION:
Trunk and Distribution SydemsaQarge scale)

Collection/ Recycling
Terminal

Small

equipment

Scale

SurfaceWater

(on-farm)

Tubewells

Deep Well

w'
el-

Itkie pmly

Communal
Sysems

*cr

d. gumi ofcmf

fcnuW
excri. The pfp

&

uy. rme i. .1-

er. e.g. wIW

rd far xultQ

b_ sm

w.id.

p;

-.

tbt W.pi*o

to rr eetib.

.-

Communal
Shallow Well

21

H. Lessons of Experiencewith Alternative Institutional Arrangements

The preceding section began from the premise that the competitiveprivate market, where
it can be applied, is best able to serve the objectivesof efficiencyand accountabilityin provisionof goods
and services. This advantageis attributedto the followingfeatures of the private sector, in pursuit of
which governmentsincreasinglyseek to involveprivate companiesin infrastructure:
(i)

managerial skill-rapid and flexible decision-making,due to a clear structure of

incentives(individualreward based on performance);

(ii)

productiveefficiency-e.g., lowerproductionand deliverycosts, stemmingfrom the

motivation to make a profit. This outcome is achieved in part because private
companiesare less constrainedthan the public sector by bureaucratic or political
issues regardingprocurementand labor practices;

(iii)

dynamic efficiency-the motivationto invest in and maintain capital equipment

needed to expandand introducetechnologicalimprovements.Investmentis also less
subjectto political influences;

(iv)

accountability to customers-the motivationand flexibilityto adapt productionto

changingmarkets and to achievebetter service quality and customersatisfaction;

(v)

financialautonomy-a final virtue of the private company, particularly from the

government's perspective, is that it reduces government's financial burden for
operatingexpendituresand debt servicing.

The institutionaloptions reviewedhere involvethe private sector to varying degrees; correspondingly,

they differ in the extent to whichthey can achievethese advantages.The analysisbelow movesfrom the
institutionalform with the least involvementby the private sector("governmentdepartment")to the form
with most extensiveprivate responsibility(privateentrepreneurship).
It should be noted that most of these alternativeforms of private sector participationhave
either not beenin practice for manyyears, havenot beentested in countrieswith differentsocioeconomic
circumstances,or have not been rigorously evaluated. This section does not provide a systematicor
comprehtensivereview of all relevant experiences.The discussionhere is intendedto illustratesome of
the more typicalexamples in developingcountriesand represent the current thinkingof knowledgeable
observers, based on a rapidly evolvingset of practices.
A. Review of Experienceby Type of Arrangement
GovernmentDepartment
In manycountries, full responsibilityfor infrastructuretraditionallyresided in govenmment
departnents, althoughhiving-off thesefunctionsto autonomousor semi-autonomouspublic corporations
or public utilities became the norm in many countries years ago. In Japan, for example,
telecommunicationswas operated by governmentministry until 1952, when it was established as a
state-owned enterprise (and more recently privatized). Governmentdepartments remain the typical

22
institutionalform for sanitation(usuallyat the municipallevel), major irrigationworks, and roads and
highways(bothnationaland subnational).Eventhe nationalrailwaycontinuesto operate as a government
departmentin Bangladesh,India, China,Egypt and manyformersocialistcountries,althoughthis is now
more the exceptionthat the rule worldwide.
As argued in the previous section, there is no a priori justification for assigning all
functional responsibility for any infrastructure service to a government department. A government
departmentis generally appropriatemainly to provide overall policy guidance and sectoral planning. It
would also be typically the channel for budgetary compensationfor specific, nonremunerativepublic
service obligations,if any, performedby the sector entities.However, since this form is so commonin
the roads and highwayssector especially,and any changeposes difficultissues of transition, it is worth
consideringimprovementswithin this particular case. For limited areas in the road/highwaysector,
operation of toll concessionsand even outright private ownershipof the road infrastructureis feasible,
and these options are discussedlater below.'4 What is addressedhere is the bulk of the sector (primary,
secondary, and urban roads) where these options are not practical or desirable, and some public
productionof road servicesis expectedto continue.
For manygovernmenthighwaysdepartmnerts,reforms are neededto (i) better link revenues
and expenditures,(ii)set clear managementobjectives,such as for maintenanceto achievea givenquality
of road services, and (iii) monitorperformanceon Li,ebasis of specific indicators. The aim of reforms
wouldbe to create surrogatemarket discipline,through a clear relationshipbetweenthe supplierof road
services and the users, in order to strengthenincentivesfor good managementand improve the quality
of service (Heggie, 1991). These objectivesof reform can be met to varyingdegrees through different
approaches, including(i) decentralizationof the road managementresponsibilitiesto lower levels of
government,(ii) corporatization,by convertingthe road agencyto an autonomousenterprise, and (iii)
contracting-outof functionsto private operators.Options(ii) and (iii) are discussedin separate sections
further below.
Decentralization.Decentralizationof road administrationis justified when it contributesto
greater efficiency and accountability.The first step often involvesreclassifyingthe road network into
those servingthe nationalmarket (e.g., 2-10 percentof the total networkin most industrializedcountries)
and those of local or regional importance.The next step is to achieve an appropriate assignmentof
expenditureand administrativeresponsibilitiesas well as of resources between the central and local
agencies. Internationalexperiencesuggests that central administrationsshould retain the functions of
strategicplanningand monitoringof a minimalnetworkof roads of nationalinterest;settingof technical,
environmentaland safety standards; and coordinationof road research and training. Regarding the
subnationalagencies, there can be true devolution of authority, whereby the local agencies have full
responsibilityfor planning,programming,and executionof their respectiveroad networks.Alternatively,
limitedfunctionscan be deconcentratedor delegatedto local officeswith centralcontrolof fundingand
supervision.Devolutionmay be more effectivein achievinglocal accountability,but it requiresa higher
level of local institutionalcapacityand political supportthan the more limitedforms of decentralization
(Cellier, 1992).
The distributionof financialresources must be made consistentwith the decentralizationof
expenditureresponsibilities.If revenuesfor the road sector derive mainly from general taxationor road

14 Toll

roads rarely accountfor more than five percentof the main road networkin any country (Heggic, 1991).

23
user charges-which may be more efficientlycollected centrally-revenue sharing needs to be worked
out through intergovernmentaltransfers or tax reform at the local level. The main issue with fiscal
decentralizationis that it may make it more difficultto achieveobjectivesof redistributionof resources
from richer to poorer areas. Certain revenues,such as vehicleregistrationtaxes and taxes on petroleum
products, may be efficientlycollectedat the local or provinciallevel.
In sum, the decentralizationof road administrationappearsa very modestmeasurecompared
to more dramatic shiftingof responsibilitiesfrom governmenttowards the private sector, but even this
measure needs to be carefully planned in view of the extremely limited supply of institutionaland
managerialcapacityin manydevelopingcountries. It can be concludedhere that effectivedevolutionof
responsibility,based on an appropriateclassificationof the road network and supported by adequate
assignmentof revenues,mayachieveimprovedefficiencyand accountabilityof performanceas contrasted
with the monolithiccentral road agency. However, corporatizationof the agency, and contracting-out
of functionsto the private sector, provide additionalscope for improvingincentivesand outcomes.
In other infrastructuresectors as well, decentralizationcan be an importantoptionto make
management more accountable to users, especially where the population affected is limited to a
geographicregion. Decentralizationcan also be a first step towards private sector involvementsuch as
throughcontracting-out,and is mosteffectivewhencombinedwithother measuresto improvedefficiency
such as through commercialization(discussed below). In the railway sector, Brazil has shifted its
commuterrail services .-, R.1.k
and S,o Paulojfrom federal 'o state control; in Argentinaa number of
inter-city passenger lines are being transf ired to the states, whiCliare better abie to assess local
demand.'5 Likewise, the national water company in Bulgaria is to be transferred to local government
ownership, through formation of regionaljoint stock companiesof which the municipalitieswill own
shares.

PublicUtility'6
This type of organization is ubiquitous in both developed and developing countries for
telecommunications,power generation and transmission, railways, ports, water supply and sewage
treatment. Many countries have begun to reform these entities to achieve greater efficiency and
accountability,through commercialization (wherebyagencies are given separate budgets and financial
autonomy based on tariff revenues, and the managerial autonomy to operate as a business); and
corporatization(whereby agencies are transformed into a legal entity subject to company law, with
formal separationof ownershipand managementresponsibilities,e.g. through a board of directors or
other body). As a publicly-owned entity, some constraints on the autonomy of a parastatal may be
legitimate, such as rules on employment, wage determination,and procurement. However, often
parastatalsare subjectedto the same political pressures facing the governmentadministrationand are
prevented,for example,from disconnectingcustomersfor nonpaymentor from payingmarket wagesfor
15 Galenson, 1993.

Public utility is used synonymously with 'public enterprise" or "parastatal" here. These entities are defined here as having
majority public ownership or control (as exercised through voting rights). Joint ventures where the private sector owns and
controls a minority share may provide positive incentiveAfor efficiency as well as additional capital, but these are not considered
here as qualitatively different from public enterprises. Cases where the privatc sector hau majority ownership and control are
treated under the section below as 'entrepreneurship'. There may also be cases of private majority ownership with minority
control, and private minority ownership with majority control (occasionally used as a transition to fuller privatization).
16

24
their staff. Public utilities therefore often function like arms of the government and their boards of
directors exist only in form.
Commercialization.Althoughcommercializationis a limitedreform in the presentscheme,
its importanceshould not be underestimated.It is a prerequisiteto making managersaccountablefor the
use of resources; it is also a necessarycondition for improved efficiency in response to competitive
pressures. In many activities carried out by governmentdepartmentsor traditionalparastatals, such as
solid waste collection,it is typicalto find officialsalmosttotally uninformedaboutthe structure of costs
(Cointreau-Levine, 1992). Explicit cost-accounting also helps to identify activities which are
nonremunerativeand, if they remainjustified as essentialpublic services,should be financedby specific
subsidies. In the United States, the introductionof commercialperformance-basedmanagementof the
road authorityhas been found to reduce road maintenancecosts by 5-15 percent through more effective
work planningand higher labor productivity(Heggie, 1991).
One dimensionof commercializationof railwayshas been to break tasks into profit centers
or 'lines of business"correspondingto specificmarkets.For the typicalgovernmentdepartment,not only
is an income statementa novel managementtool, but so is a balance sheet. The characteristiclack of
orientationof public officials to the conceptof asset managementmay help to explain the widespread
problem with inadequatemaintenanceof the infrastructurestock.
Corporatization. The Bank has a long experience in supporting government efforts to
corporatizepublic enterprises,along with other reforms to enforce market-like pressuresby cutting off
access to subsidies, and enforcingbankruptcyand liquidationof enterprisesthat prove nonviableunder
these conditions.In very many cases, such public enterprisereforms have been of limited success, and
the gains achievedhavebeenephemeral.The pervasiveproblem is that wheregovernmentshavedominant
ownershipand controlof enterprises,political interferencein managementwill sooner or later reemerge
(Kikery, Nellis, and Shirley, 1992).Therefore, where feasibleand appropriate(accordingto the decision
criteria outlinedearlier), it is often preferablefor public enterprisesto be convertedto private ownership
through divestitureof shares or sale of assets.
Despitethese shortcomings,corporatizationrepresentsa considerableadvanceover current
practicein some areasof infrastructure-such as managementof roads, major irrigationworks, and solid
waste transfer and disposal.In fact, for those segmentsof infrastructurewhere there are high sunk costs
(specificallyfor the network of trunk facilitiesof major highways,power transmissiongrid, urban tram
and general rail lines, and water and sewerage pipelines),the most appropriatearrangementmay be for
ownership to be vested in a public corporation which will take a particular interest in ensuring a
reasonablereturn on these assets.
Sincepublic enterpriseswill remain a dominantorganizationalform for many infrastructure
activitiesat least for a transitionalperiod, methodsof improvingtheir performancecontinueto be needed.
Various ways of restructuring the activities of the enterprise, such as by "unbundling" the fixed
infrastructurefrom the more contestableoperations, are possible. The Swedishrailways, for example,
separatestrack ownershipfrom operationsto facilitatecommercialoperation (Galensonand Thompson
1992) discuss this and related options in practice for the railway sector in a number of countries.

25
Additional approaches covered in later sections below include management contracting, service
contracting, and leasing."
Although it is still relatively rare in developing countries, the conversion of road
departments to public utility corporations is attracting increasing interest as a way of improving
3 In practicallyall countries, the sector has an available
performance,particularlyof road maintenance."
source of financingthrough road and fuel taxes. The reform entails identifyinga portion of these taxes
as user charges-the "road tariff-which would be channeled directly to a fund managed by a
corporationwith responsibilityfor road expenditures.The road utility would have to be answerableto
trustees as "owners", representing both road users and the government. Such a system exists in the
NationalRoadsBoardof NewZealand,under whichroad expendituresare budgetedbased on assessments
of traffic-relatedcosts; user chargesare then calculatedto reflectthe wear and tear createdby different
types of vehicles; and the Board manages its budget autonomously,but with accountabilityto users
(Galenson,1990).
In the municipal solid waste subsector (MISW),the activities of transferring wastes to
disposal points and disposal itself exhibit economiesof scale and are most efficiently managed by a
metropolitanarea-wide or regional authority. Cities in a numberof countriesare using quite effectively
a public servicecorporation-typeof authorityto performthe transfer/disposalfunctions(Bartone, 1991).
In such corporations,the local governmentsinvolvedmay serve as owners(partiallyor wholly);the entity
could also be privatelyowned and regulatedby the local governments,or publicly owned but operated
as a private concession(as discussedfurther below).
Significantchangesin orientationare requiredin the transitionof a governmentdepartment
to an effective corporate entity, as illustratedin the case of the National Water Supply and Drainage
Boardof Sri Lanka (see Box 2). An evaluationconductedin 1988of five of the most successfulmunicipal
and regionalwater and sewerageutility companiesin Latin America, which comparedtheir performance
to that of Spanishand French utilities, concludedthat the most successfulcompaniesin both groupshad
the following features: (i) high quality managers and stability in mid-managementand professional
positions; (ii) financial strength, including reasonabletariffs which cover at minimumoperationaland
maintenancerequirements;(iii) an emphasison maintaininggood customerrelations; (iv) in some of the
companies, successful use of private contractors to perform O&M function, participation of private
capital, and the use of effectivecost accountingsystems.It was also notedthat very clear delineationof
roles of the government, the board of directors, and the managementcontributedto top performance
(Yepes, 1990).
Other approacheswhich have been used to help public enterprises in developingcountries
improve performance includetwinning and contract plans. Twinninginvolvesa horizontal, sustained
cooperationbetweena well-run, mature companyand a less experiencedor poorly performingcompany.
Twinninghas been done in water and power, for examplebetween NEA, the nationalelectrical utility

17 These three alternatives (in particular service contracting) may also be used by government departments, but they are
discussed here in relation to parastatals.

la The UN Economic Commission for Latin America and the Caribbean (ECLAC) has a research project supporting the
development of this institutional arrangement (Schliesser, A., 7he Reform of Road Network Conservation. Basic Issues to be
Treated, Conference on Commercialization of Public Transport Services in Developing Countries. Buenos Aires, 1991.)

26

Box 2 The National Water Supply and Drainage Board of Sri Lanka-Improving the Performance of a
Corporatized PubiUtUtility...
In 1975. the Departmentof Water Supply and Drainage, within the Ministryof Local Government,
Housingand Construction,was almost entirely devotedto perform capital constructionprojects, mainly to create new
capacity.From 1975, it was transformedinto an autonomousbody, the NWSDBwith responsibilityfor developingand
operating water supply and piped sewerage. By 1984, the company was functioningreasonably well in terns of
constructionof new schemesbut had been demonstrablyunableto shift its role from a focus on capital projectsto O&M
and consumerbilling. The USAIDIWASHproject for the institutionaldevelopmentof the companybegan in 1985with
the gim of ssistingit to radicallyreform its roecto be consistentwith its corporatestatus. Over the next 6-7 years of
intensiveconsultantinput,the projectfocusedon developingcorporateplanning,regionaldecentralizationand delegation,
improvedO&M(reductionin unaccountedfor water, improvementin water quality, preventivemaintenance),improved
financialmanagement(performancebudgeting,tariff reform, cost control),managementimprovement(staffoptimization
and employeeperformiance
evaluations),and customer relations.Indicatorsof performance,includingcollectionsratio,
billing lag, consumer complaints, unaccounted-forwater, and the rehabilitationshare of the capital budget were
considerablyhigher by the end of the project.
The NWSDBinstitutionaldevelopmentprojectwasconsidereda successin shiftingthe organizationfrom
engineering-orientedto an O&M and conmmercial
orientation.Key to this successwas the intensiveconsultantinput, but
sustainabilityis judged to be high because of the decentralizationof responsibilitiesto regional centers, and the
establishmentof an effectivecorporatemanagementcapabilityin the institution.It is interestingtbat the project did not
involve contracting-outfunctionsto the private sector. The evaluationreport notes that external politicalintervention;
continuesto exist; the cornpanywas forced to accept a 45 percent government-mandatedsalary increase in 1989, for
example;the governmentapprovesand financesall investments,and there are politicalpressures for hiring. To insulate
the companyfrom such negativeexternal forces,the projectstressedthe developmentof cost consciousness,an in-housc
planningcapabilitywith clear projectselectioncriteria, emnployce
performanceevaluations,and a corporateidentity,
Sources: Executive Sunmaiy FmalReport on Institutional Development of the NWSDB, August 1991. Engineering-Science

Inc., Pasadena CA., in colUaborahion

with ResourcesDevelopmentConsultants,LTD, Sri Lanka.

of Nepal and Electricity de France (EDF). The two-year relationshipbegun in 1989 helped NEA
managementto understandthe reasons for poor performancein the past, and to clarify its role relative
to that of the government. However, the arrangement was unable to affect the continued extensive
interventionof the governmentin NEA affairs. A weaknessof twinningis that the senior utilityis limited
in its ability to influencedecision-makingby the junior partner.'9
Contractplans (also called"performancecontracts")are negotiatedagreementsbetweenthe
government,acting as the owner of a public enterprise, and the enterprisemanagers. Contractplans also
help to clarify goals of the utility and its role relative to those of the government.They usually involve
a promiseof increase managerialautonomyfor the enterprise in exchangefor its fulfillmentof agreed
pertormance targets (Nellis, 1988). For example, in the above case of NEA in Nepal, the twinning
arrangementwas followedby a performancecontract to formalizethe agreementsreached betweenthe
utility and the governmentregarding their relationship.Contract plans under various nameshave been
used for the railway in Senegal, Mexico, and Kenya, with very limited results.0 Contract plans can be

19From EDFInternational,'Seminaron Power UtilityManagementby PerformanceContracting', WorldBank, May 19, 1992.

2

Galensonand Thompson,1992.

27
helpfulin identifyingand monitoringof performancetargets, and workbest for companieswhich already
operate commerciallywith sound financial and reporting procedures. A major shortcomingof these
agreementsis that the enterprisecannothold the authoritiesaccountableto enforcethe contract,and many
governmentshave reneged on their commitmentsunder these agreementswhen political or economic
21
conditionschange.
Service Contracting
Service contracting involves "contracting-out' of specific operations and maintenance
activitiesto the private sector, usuallyfor a periodof a few years. Withthis approachthe publicprovider
(governmentdepartmentor public enterprise) sets the performance criteria for the activity, evaluates
bidders, supervisesthe contractor(s),and pays an agreedfee for the services, which may be based on a
lump-sum, unit costs, or other basis. To achievegreater efficiencygains from contracting-out,contracts
should be awardedthrough competitivetendering;private bids may even be compared againstthose of
the public agency.Some countriesretain a forceaccountcapacityto competewiththe private contractors,
or to provide a residualmeans of performingessentialfunctionsshouldthe contractorfail, for example
due to a labor dispute.However,sincepublicagenciesnormallyhave poor cost accountingwhichrenders
it difficultto comparetheir service coststo those of an outsidecontractor,it is importantthat competition
for contracts between the force account and private contractors be based on a full and comparable
assessmentof costs.
Activities such as ticketing, cleaning, food catering can be readily contracted-out for
railways,as done in Pakistan,and locomotiverepair and maintenanceare carried out privatelyin Kenya,
of standpipes,meter
Suday, and Senegal.In the watersupply/seweragesector,the operationlmaintenance
reading, billing and collectionsare also obviouscandidates,and such service contracts have been used
in Chile sincethe 1970swith good results in staff productivityand cost containment.Where there is need
for particularlytight coordinationand quality control, such as in maintenanceof track infrastructure,
contracting-outis still possible but requiresespeciallycareful oversightby the public authority.
Sometimespracticallyall of the corefunctionsof publicagenciescanbe contracted-out,with
only a basic staff left in the agencyto award and monitor the contracts. This is the nature of AGETIP,
the nonprofitcontractingagencyfor small scale infrastructureprojects financedby World Bank projects
in Senegaland several other West African countries(see Box 3). Anotherexample is the public water
companyin Santiago, Chile (EMOS),which in 1977 introducedcontracting-outas part of a strategy to
reduce its own staff and use their servicesas private contractors.The approachhas contributedto EMOS
achievingone of the higheststaff productivityrates in this sector in Latin America, even when the labor
content of the contractsis taken into account (Triche, 1990).
Contracting-outof road maintenancehas been an establishedpractice for manyyears in a
relatively few countries (includingBrazil, Kenya, and Malaysiaamong developingcountries), and has
been introducedmore recently in many others (includingAlgeria, Chile, and Pakistan).A 1991study of
this arrangement in these as well as several developed countries made some interestingobservations.
After experiencehad been gained with contract maintenance,there was a general perception in the

21 However, the governments can hold the enterprise accountable: in the Gambia, three public enterprise managers were
dismissed due to failure to achieve the goals stipulated in contract plans (7he Reform of Public Secior Management.- Lessonsfrom
Experience, World Bank Country Economics Departmcnt, Policy and Research Series No. 18, 1991.)

28

.. .. ..

::~::::

tract MU agementof Publi Work:

;70~l ~; ~-l
: ;it
-li*AGETIP
Th
modelof contractmanagement.
firstusedin the contextof aJBankprojectin Senegal,
wasV
:-developedfto improvethe efficiencyfofpublicworks projects by permiittingthe applicationof privatesector management
p-to
practicf
the publ:c;sector, transferringdie executionof sub-projectsfrom inefficientpublicagcnciesto dynamicsmall.
;0firmsfintlheprivaitcsector,$and libtig
the managersof projects from the cumbersomne
red tape of public institutions.
AGETW:
(Aga:i;icfrte
Eection
of
Work
in
ithe
Public
Interest)
is
a
private,
not-for-profit
company-an NG0. Its;
..' ;.; E 's frs.ea
0
r oo peato,
n t GT i.e.ga
i' ..... '.;.ii.;.i''tt
..... ....
....
mlloas municipalities
ofawoks throughc119nsubprolects
missitE
on.eued.
is :liun4beskegeneral
contractingcommfissions
from public entitiessuch
or ministries,It hires:
*coaaultantstopreparedesigns andfbiddingdocumcntsand to superviseworks; it issuescaLl4forEbids, evaluatesthemand?
sig Ib contracts
itz evaluates
progress,i
paysthiecontractors,andundertakes
the final ecpion ofthe works.Actvities:
....of teagenyreegulated
throughits Manualof Procedures.

~(z~
hETIP
lsoPaiysmaillor
ionfers
indiretbnef
othreating
e eeconom,byprevatin
dmndt
rfitorndpeny-ricesGof
loca
....
using.7 *'-'-'.''.-"-'.-'-:0'''--.'400
oediumsized
contractors
a1Pulcal
almoust2000person-years
of~6
ejmploymentA number
'':;
f 06 jcand
of fatorsex.plain.its.success..first,theprvate sectormanagementteamactivelyseeks outefflcieney;for example,it takes
pride paying
t:
onctors in 10 days, rather than the 30 days allowed in the manual or the monthstaken by publio
t...
.tie.s..Sc..ceit
ontracts engineering consultantsand contractorson an as-needed"basis, its overpead isv
-iiextrly low. 'Third its privtedlegal status absolves it: frome the cumbersomebureaa
procedures of the public

c..:-o'ntrcing-s:d
consulting
industries,thus stsimulating
their development.AGBTUP
currentlyhas 680 local contractorsand:!
i-:6(l ai dconsUkta0son
0itsotoster. The successiofAGETIPhas alsc
espurred other agenciesto improve
their:ownIt
performaxce;the-Minister:of PubliicWorkcsis consilderingsettingup an AGETIPwithin;his ministz7,:and the mayor of:
i.':.SDakahas:-used
AGEIl?to execueprojects financedfrom his own budget. 0:
:::::::::
::::
:
:::i
t;i'li "-t-'t
.$iTi:.
i;$G'ET;
'I'tP's
successS
hs led to the craion of similaragencies underWorld Bankprtjecb in Niger, Mali,7E
-'Beami, tand Burkinaaso,;:Mauritaniiais aboutf to: create such an agency as: well.: One component of the propo:serd?:
.......
Maurit projectwill disseminateinformationto localcommunitiesinworder
to increasetheir sense of responsibility
.t.o.w project work. An obje e is to improvecommunicationbetweenthe grassrootsand local authorties so that the;
.fo:rmer are consutedaprojectsthat:affcqt their neghborhoodbeforeatheproects arresented for funding. Thisvis::
pt'0:t
articularySipantlit-h Atheicaseof :garbagecollection:and sewer cleaning,:whSeregrassroots:participationcan make a
:ed:
:::
ifdifence in terms of ipublic:hygierMe:u:fPT.i

countriessurveyedthatit is morecost-effectivethanforceaccount,althoughdirectcomparisons
of costs
betweenthe approachesweredifficultandincompletebecauseof differentmethodsof execution,quality
standards,andaccounting
practices.Itwas alsogenerallyacceptedin thecountriessurveyedthatthe force
accountworkgroupsbecamemoreefficientwhen they had to competewith privatecontractors.It is
interestingthat all contractorssurveyedexpressedthe need to expandthe scope and durationof the
maintenancecontractsto allow for capitalization
and for the acquisitionof specializedequipment;this
findingsuggestsa tendencyfor service contractingto expandgraduallytowardscontractsinvolving
investment(concessions).ln countrieswheretherehadbeen a successfultransitionfrom force account
to contracting,a close workingrelationshipanddetailedinformationexchangesamongthe roadagency
andthe contractingindustryhadbeen importantfeatures(Miqueland Condron,1991).
Collectionof municipalsolid waste by contractis becoming increasinglycommonin
developingcountries,includingsome of the poorestcountriessuch as Haiti. The impetusfor seeking
privatesector involvementin this activity is obvious: expenditureon MSW services absorbs20-50
percentof totalmunicipalrevenuesin developingcountries,yet on averageonly 50-70 percentof solid

29
wastesare collected.Studiesof privatesector involvement-mainlythroughservicecontracting-in MSW
in both developed and developing countries have concludedthat private collection is generally more
efficientthan public (Cointreau-Levine,1992; Bartoneet al, 1991).The private contractorshave lower
wages and benefit costs, younger (more productive) crews, and face fewer civil service-or
union-induced restrictionson hours of work and overtime. Private sector crews also have newer and
more standardizedvehicle fleets, allowing for greater productivity, and higher vehicle availability.
However, in this subsector, as well as in road maintenance,it is often difficult to compare exactly the
cost-effectivenessof private versus public agentsbecause of the general absenceof cost accountingby
the municipalsolid waste department.
It is importantthat service contractsbe of sufficientlengthto permit the contractorsto fully
amortizeany equipmentpurchased,such as collectionvehicles(aboutfour years). In developedcountries
where large marketsexist for resaleand leasingof capitalequipment,this argumentfor lengthof contract
is less of a concern.
Management Contracts
Managementcontractingextendsthe responsibilityof the private sector beyond individual
servicefunctionsto encompassa broad scope of operationsand maintenance,usuallyfor a period of 3-5
years. In cases wherethe contractorreceivesa set fee for servicesrendered, the arrangementis essentially
no different from technical assistance.In the concept of managementcontractingaddressed here, the
contractorobtains at least some of its compensationas a functionof the company's performance, and
therefore shares some of the commercial risk of the enterprise. For example, in France, where
managementcontractsare commonin the WSSsector, the incentivesfor productivityimprovementsare
provided by linking the contractor's payment to indicatorssuch as reduction in leakages or number of
connections. Box 4 describes how technical assistance evolved into a management contract in
Guinea-Bissau.
If compensationis linked to performance, it is necessary for the contractor to be given
autonomy in day-to-day managementdecisions. In many developing countries, however, the private
contractor cannot control key functionsthat affect productivityand service quality (such as staffing,
procurement,or publicly-providedworkingcapital), and thus cannotbe held fully accountablefor overall
results and evaluationof performance is difficult. In these circumstances,managementcontractorsare
often compensatedon the basis of a fixed fee. Greaterefficiencygains canbe obtainedfrom management
contractingwhen significantautonomyis granted and the compensationcan be based at least in part on
performance. The contractsshould also incorporatesome incentivesfor training of the internalagency
staff, although this has proven difficultbecause of the short-term nature of managementcontracting.
Given these problematicaspects,managementcontractscan be very useful as interim arrangementsthat
allowprivate firms and public agenciesto gain experiencewith partnershipsprior to engagingin more
comprehensivecontracts such as leasing or concessions (describedbelow), or while reforms of the
regulatory framework for the sector are being developed. This has been the trend of management
contractsin France, where they have tended to become more comprehensiveover time.

30

Box 4 ''TheTraztion tron Technical

wAssistance to Management Contracting:The National Electrical and
(EAGtH)hi Guinea-Bissa-d-n
t'WaterUtiity
l
.....
t! ;0;
; '7-Tn 1986,iAthe faoeof EAGB'anear bankruptcyand disastroujoperationalperformance,the government
of Guinea-Bissau t.trequestedtcchnicalassistancefrom Electicity de France (EDF), a Frenchpublic enterprise.By
.I989,
technicalassistancehad achievedddramaticimprovementin the,utilit's operationalefficiency,but its financial.....
thios
ncrissremained. EAGCBandEDd then eterd into atwo-year managementcontract, which was renewed in 1991....
Ac;ordingto the terms0of the contract,EDFtook overthe operationof t utilit, withEDFs compensationbasedpartly
on fixed fees (paid: bythe FrenchMinistry of Cooperadion),and patly on a share of the revenue-basedprofits of thc
utility the uizc of this profit share dependi on the rtalizition of performancetargets, Thc pcrformanceindicatod
n used
includeth tiof powerbilledto the net power generatedand delivered,and the bill collectionratio. The valiei of these;
iMdicat
verifiedby an independent
auditor.
- -dt;i0g:t:10e ' Between1987and 1991,EAGBachievedstronginiprovemnentS
in powergeneration,operationalefficiency
andfinancialcefficiency. Accordingto EDP's assessment,there are drawbacksof the management ontractapproach.
:Ccluding In particularthe likelihoodof conflictsbetwveen
in
the externalmanagersand the local utility, the limite incentive
tiet outside
.g05'lf' contrctors to provide traning toithe local staff if performanceiusmeasuredby shortterm results;:andg:00
...- potenial confusionas to what will happento the companywhen the contrat expires. In the EAGBcase, EDF
.thatnots
.... thmain problemsof the utility.s.te.. om shortagesof physicalcapacity ew investmentwasno part of the.manageent
oontract),.and a9continuationof paymentarrears rom governmentand parastata customersof the utility.al ici
e: From seminarby ED iF
em Uto
tili ManagementabyPerformanceContractingt, presentedto
Worl
0 Bank
on
Ma 19.; 1992.0;;:
Internationalon "Power --- ;:A-iE
;; :i:: 00d- tS00it-- ySi:::00::;

Lease Contractsya
Leasinginvolvesa privatecontractorpayingthe publicowner for exclusiverights to operate
facilities(withoutresponsibilityfor majorinvestments),and bearingfull commercialrisks. A lease
contractaccordsan exclusiveright(sometimescalleda franchiseor license)to the streamof revenues
from providinga service.Leasinghas been used for decadesin urbanwater supplyand seweragein
FranceandSpain(wherethe arrangement
is knownas "affermage"),andhas also beenused elsewhere
in power, ports, urbantransport,railways, and solid waste collectionand disposal. In the case of
"landlordports",for example,the governmentowns the landand infrastructure,
and the privatesector
owns andoperatesthe superstructure.
In MSW,the lesseeobtainsexclusiverightsto the solid wastein a given residentialzone,
whichis the "assetowned"bythemunicipality;
thisexclusivityenablestheoperatorto captureeconomies
of contiguitywithinthis zone. In thetransportsubsectorssuch as urbantransit,leasingis moreoften
called "competitivetendering"for the rightsto operatea particularroute. This approach,which is a
middlegroundbetweenfull deregulation
andpurelypublicoperation,maybe particularly
relevantwhere

z This term may also be used with referenceto an operator's rentingor leasing particularequipmentor vehicles, in order to
reduce the costs and liabilityof owning *uchequipment.This sense is used, for example, in the case of municipalitieswhich
leasesolid wastecollectiontruccs fromthe privatesector in SantaCruz, Boliviaand Metro-Manilabecauseof the governments'
constraintson capitalinvestment(Cointreau-Levine,1992).Suchlimited-purposecontractingis akinfservice
to the
contractingt
discussedearlier.
3

Cointreau-Levine,1991, op. cit.

31

there are concernsto avoid excessivetraffic congestionand air pollution as a result of many operators;
or to provide socially desirable but unprofitableservices. In the U.K., the right to obtain subsidiesfor
certainnonprofitableroutes as a socialserviceis awardedto the lowest bidder. Competitivetenderingcan
be used to grant the rights to provide some or all services in the market, either to replace the public
service or compete with it. The State Railway of Thailand introduced competitivetendering for the
operation of certain passenger lines in 1985. As of 1990, the lease was found to have succeeded in
attractingroad users to the railway whilegeneratinga profit (Levyand Menendez, 1990).
In lease contracts, the public owner (lessor)remains responsiblefor fixed investmentsand
debt servicing. The contractor (lessee) normally must finance working capital and replacement of
short-lived assets, such as small-sizedpipes in the case of water supply. Accordingly,the duration of
the contract is usuallybetween6-10 years, correspondingto the amortizationperiod of such works. The
contractorusuallycollectsthe tariff revenuedirectlyand returns an agreed portionto the public authority
as a rental or license fee. The profit for the contractor is the difference between the gross revenues
collected, and the sum of operating costs and this fee. Any savings from efficiencyimprovementsare
thereforeretainedby the contractor.The leasecontractshouldspecifythe maintenancerequiredto protect
the condition of facilities during the lease period; the performance indicators to be used for judging
quality of service; procedures for enforcement,penalties for nonperformance,and means of dispute
resolution.
The experienceswith lease contractsin the watersupply sector in Guinea, and in power in
Cote d'Ivoire, are describedin Box 5. Cote d'Ivoire has also had leasingfor urban water supply in the
past, and a lease for water, sewerage,and electricityis plannedin The Gambiawith the assistanceof the
EEC.
Concessions
A concessioncontractincorporatesall the featuresof a lease, but with the contractorhaving
the additionalresponsibilityfor financing certain specified extensionsor replacementsto fixed assets.
BOTs are concessionsestablishedfor entirely new investments,and these are discussed further below.
Concessionarrangementsexist in power, watersupplyand treatment,solid waste disposaland treatment,
ports, railways, urban metro systems, toll roads, and telecommunications.They are not widespreadin
any of these sectors, however, in developingcountries. The concessionapproachhas been used more
widelyfor specialistports, such as grain and ore terminals,than for commonuser facilities,but examnples
of the latter exist in Thailand and Taiwan (Scurfield, 1992). Some of the longest experience with
concessionshas been in the municipalwater sector in France (see Box 6) and in Spain (Barcelona).
In a concession,investmentplans and implementationare subjectto review by the authority
issuing the contract. The assets revert to the public owner upon completionof the concession. The
contractor's compensationis based on tariffs, which are also determinedaccordingto agreementset out
in the concessioncontract. The tariff revenueshould be sufficientto cover the operationalexpensesas
well as debt service and depreciationon the concession'sinvestments.Concessionsnormally last 15-30
years, dependingon the life of the investments,and they are often renewed. Concessions,like leases,
require appropriatedisincentivesfor contractorsto run down the assets towards the end of the contract
period, althoughthis wouldbe less a concernif the incumbentintendedto competefor contractrenewal.

32

'Box
--'5 Lease Contracts ' t

Urban Public Utites in West A fi .* ...

wate1develop
'-.Suv.lv
Water
in Guinea. Prior to 1989, The Republicf Guineahad one of the least
soctorsin West Africa.: At. that time, :the sector was rcstrutured: ownership of the: iban water s Ipply
::0Z
finfastrcure aind :rexponsibilitykforsector planning
investmentwere transferred to a new autonuma
ter
authority, SONEG; Canewcompany, SEEG, was createdto operateadi maintainthe facilities.SEEG is a.joint
.eture
of which49 percent ii owned by the:government and:51 percent by a&:foreign
pnrvate consortium. Thestrength othi i
: : g :Guinean
amrangement
lies in the clarity:of responsibilitiesand incentives.:Underthe ten-year Iase contract sge wih
,SONEG.SEEGoperatesand maintainsthc systeat its own risk, with its:remunerationbasedoniuser:chrgaactu,ally.colectd andbnewconnections.SEEGbenefits from improvementsit achievesin the collectionratio, andin retduction.o
:operiting costsand unaccountedfor Water.:SONEGhas incentivesto seekadequatetariffsand to investpruddntlyabased
on realisticdemand forecasts,since it has ultimateresponsibilityfor capital financing.To:date the collection:ratio as
increaaed dramatically,from 20:to. 70 Opercent,
and technical efficiency and coverage have6imnprov
ed.: Tarifi have
increased&ddm
60 GFtm3before the lease contractto 420 GFlm3 (about US$.45)in 1992, and are.expectedto reachdth
lli cost recoverylevel by 1998, In the interim,ta World Bank credit is financingthe foreignexchangeicostsof th.....
oi:perationan a decliningbasis, and the governmenthas assumeddebt serviceon a decliningbasis. The arangement .ha..
beenampe red, however, by delays unprocurementof equipmentby the state cterprise SONEG,
has
.. .te
.wh.ich
::'SiSEEG's
ability to inprove th:equalityof serviceand its financialperfornance.
...
:....
.-

dupply

;l-0Source:Thlt-1cma
Trche,jPrivatePariciation ;ii the Deliveryof Guiiiea's0WaterSuipplyServices,;Wor'kingPape.r.No ..'
VPS 477,0World BanikINUWS:,:1990;:
Wt;
and PSP hn UrbanWater Supply. Lrrues,impliatons. end &0Enaples,;
INUS;..

:f-:Note,
August
:1992.

:: g:

::g,::

g::g:g:tgiiE::i;

:::i

0::-

Egg
:75i:igggi:
:.;:::: i:: : Powver
sectorin Cote d'lvou_.:The performanceof EEICI,the parastatalresponsiblefor the electricpo,wer..a.i:-tector
in Coted'Ivoire, deterioratedduring the 1980st.In 1990a major restructuringwasiundertakenwheirebiyoperations.'.-.
::.Dtand
maintenancefunctionsfor geaeration, transmission,and distributionwere transferredunder a leasingarrarigement.:,'
tlat newjoint venture,tthe Compagnie
to;;;|
Ivoirieune;d'Electnicity(CIE). Fifty-onepercent of CIE:isfownedby a consortium i
'of twoPreich companies,and 49 percentby lvorian investors
. EECI continuesto own the facilities-and
berresponsibiity:-'-:
fl;'-0or
investments,;sectorpolicy tad plninng. fIn the irsjt IS8monthsof the newarr,angement,
efficiecynd service has.g,}
im:';:prove
marlcedlyi
(CUEhas improvedthe: collection rati firom60 to :90:pexcen.t;
maintenanc has incease; dura*iono000
power
;f outagesreduced;businessoperatos computeied; andoperatingisubsidies
havebeen elimninated).
An important
i;0factorin this successis Mthe
creationlof CIE!as a new operatinlgentity, whichestablisheda 'formal:breakwith the pastevenl1
0though C took0over someof.the staff.of.EE.CI.Ir addition, C.E..obtainsconsiderabletechnicalsassistance romits.
tl;.tSource:EDE International,Tiower UtilityCManagement
by Performanc Contrcting"0.Seinar

pre-sentedto World Bank

A concession for water supply services in Cote d'Ivoire waS arranged in 1987,
following25 years of a lease contract. The lease had achievedgood results in improvingservice and
internal

efficiency

of the operating

company,

SODECI,

a mixed

French/Ivorian

enterprise;

however,

financial troubles mounted in the 1980s because of policies enforced by the government regarding
investment

and tariffs,

for which it retained responsibility.

Under the new concession arrangement,
the
SODECI
became both operator
and investor,
with responsibility
for all new urban water supply
investments
in the country. The company receives no operating subsidies and all new investments are

self-financed.Althoughthe concessioncontractwas not subjectto competition,SODECI's operatingfees

were reduced substantiallyduring the negotiations.The company's operating costs are comparableto
those of many water utilities in West Africa, while its service quality iS far better than most. Private
Ivoirianinterestsnow own a majorityof SODECI's shares, and the companyhas succeededin reducing
expatriate

staff while expanding

operations.

33
Amongrecent examplesof this arrangementin developingcountries,a concessionfor water
supply and sewerage in Buenos Aires, Argentina, was awarded in late 1992; and as part of the
restructuring of the railway in Argentina, six separate cargo lines were created and awarded by
concessionagreements.Initialresults form the railway that has been operatingprivately for the longest
time are quite positive. Operatingon a line which had had little or no traffic under the nationalrailway,
the new company (FEPSA) has put a major emphasis on marketing and worker motivation and is
competingeffectivelywith trucks and improvinglocomotiveavailability(Galenson, 1993). The Buenos
Aires subwaywas also put out to a concessionawarded on the basis of the least subsidy required to
operate and invest in the system. A recent unsuccessfulattemptto arrange a concessionfor water supply
in Caracas provides some importantlessonson the need for: (i) making available informationon the
quality of assets and costs of operationprior to requestingbids; (ii) high-level politicalsupport for the
reform; (iii) adequatetariffs and projectedrevenue levels; and (iv) adequatetreatment of political and
economicrisks, particularly exchangerate risks to the private participants.
Build-Operate-Transfer (BOT) schemes.' The experiences with successfullynegotiated
BOT schemesare relatively few in developingcountries,and are reviewed elsewhere(Augenblickand
Custer, 1990; Besant-Jones, 1990; and Israel, 1992).It is increasinglyrecognizedthat such schemesare
extremely complicatedand time-consuming to arrange, and few have become operational to date.
Experts advise that if countriescan finance new infrastructurein a more traditional way, as a turnkey
construction project financed by government borrowing, the time saved and greater certainty of
completionmay warrant the traditionalapproach. Of course, countrieswhich cannot borrow sufficient
investmentcapital directly are more interestedin the BOT schemes, which have "limited recourse" to
governmentfinance. However, in manycases, such arrangementshave involvedconsiderableexplicitor
implicitguaranteesfrom governments,of traffic volumes, revenuelevels, input prices, etc., so that the
governmentsare left with considerablecontingentliabilitiesafter all.
TheBOT-typearrangementswhichhavebecomeoperationalin developingcountriesconcern
mainlypower and transportprojects in Asia. In the water supplysector, a coupleof BOTs have beenset
up in Malaysiaand one in Indonesia,for source supplyor treatment. A BOT for waste water treatment
and reuse by industry has been successfullyimplementedin Vallejo, Mexico. Financed entirelyby the
participatingindustries,it is operatedunder a renewableconcessionfrom the local water authority,which
managesthe distributionsystem linking the industriesto the treatmentplant.
It is importantto note that even in cases where BOT-type arrangementsare successfully
concluded, they do not absolvegovernmentauthoritiesin the sector from the basic responsibilitiesof
sector planning,policy-making, and investigationof project feasibilityas well as all the tasks involved
in the contract-lettingprocess. The public authority or sponsoringministry has to develop the detailed
specificationsof the tender documentsin order to ensure that bids are comparableand serve the public
interests.The public authority shouldalso ensure itself that a proposedBOT schemeis financiallyviable
(and on what terms, i.e., possiblegovernmentguarantees),and investigatealternativeoptions whichmay
be more attractiveeconomically.
Toll roads. Many developingcountriesare lookingto the optionof toll road concessionsas
a way of generatingprivate fundingfor road constructionand operation,or simplyto movethe operating
costsof roads out of governmentbudgets.Toll roads can operate in a well-definedniche of a country's

24

This sectionis meant to refer to other variants of BOT schemesas well, such as BOOT (build-own-operate-transfer).

34
..... . . . ... . .. ....

Boxj6 0Two:
Models of.Private Sector Particivpationa Water Supply: U.K. and France

-atr supplyand sanittion servicesinEinglandand Waleswereprivatizedin 1989, when sharesin the

-'tenpubli water authoritiestwere sold to the public.:The primary objectivc was to transfer to the private sector
responsibitity
forthestaggeIngly large investmentprogramneededto make up for past neglectof assets and to bring the
U.K. syst Vp to newEC standards.
......0
0 }Prices arelregulated
througha pcoeap formulathat links water ratesto the retail price index plus a
::factor (4"K), determinedindividuallyfor each Company,that taces into account the cost of the investmentprogram. A
: ;::centrl aiuthoritymonitorsand enforoesenviromnentalstandards,a functionpreviouslycarriedout only laxlyby the water
a:000..suthoritisthemselve.
The cost to the British governmentof this privatizationwas significant:the governmentwrote off
debts totallinkg
L449billion, transferredthe existing cash balancesof the.water authoritiesto the new companies,and
:::'Sinjected
an:additionalLI .6 billioninto selectedcompanies.The regulatorymethodchosenhas highinformationcosts and
providesno cleaincentive for the companiesor the regulatorsto economizeon the costs of meetingrequired standards.
On the benefits side are improvedenforcementof environmentaland drinkingwater standardsand the promiseof more
reliableand efficientservice.
Assetownershipin Franceremainswiththe municipalities,whichcontractout operationsand maintenance
toprivatefirmsundermanagementcontracts,leasecontracts,or concessions.Underlease contracts.andconcessions,fees
are establishedthroughnegotiationor on the basis of competitivebidding, withadjustmentsfor inflationaccordingto an
agreed formula;changes in eircumstancescan trigger a renegotiation.Tariff levels are monitored by the Ministry of
Economyand Finance, and environmentalstandardsare enforcedat the:departmentlevel.
-Competition
is a keyto efficiency.Whilethere is no directcompetitionamongthe U.K.watercompanies,
they compete
pfor capital in the stock market with other investmentopportunitiesand through the use of benchmark
i:..ndicators by the regulatoryageney. In France, the water supply market of each municipalityis contestable,and the
existenceof Itwo.major contractingfirms (Lyonnaisedes Eaux and CompagnitGeneraledes Eaux)also permnitsthe use
iof benchmarkindicators.Neitherof the two arrangementsnecessarilyensuresefficientinvestment,however.The U.K.
price-cap approaehis intendedto avert over-investment,but the K factorswere set without any apparent cost-benefit
analysisof the neededinvestmentor scopefor cost minirnization.Furthermore,the approachappearsto be movingin the
directionof de facto rate--of-returnregulation,as highdividendsand large increasesin managers'salarieshaveprompted
a reviewof price caps soonerthan the planned 10 years. In France, separationof ownershipand operationscoulddistort
ii;tnvestment
incentives.One mechanismthat helpspreventthis is full cost recoverythroughwater tariffs; since both owner
and operator recove costs from the same source, they have an incentive to collaboratein making sound investment
decisions.The length of the concessionsmay also result in better asset managementin France; on the other hand, it
reduces the potential for competition..
The complexityof marketingwater and sewerageassets and lack of interest on the part of the private
sectorto assume responsibiltyfor rehabilitationand investment,alongwith the heavyregulatoryrequirements,may make
the U.K. modelunattractiveto many developingcountries.The French system, on the other hand, is already in use in
severaldevelopingcountries.The approachcan be adaptedto fit a varietyof situations.I f properlydesigned(for example,
with compensationlinked directlyto volumeof water deliveredto consumersand collectedfor), such arrangementscan
transfer all of the commercialrisk of operationand maintenanceto the private operator.Actual ownershipof the assets
may not matter very,much;what is importantin both cases is that the entity -owns" the right to a stream of revenues.
Source: ThelmaTriche and John Briscoe(INUWS),
basedon seminarson "Privatizationof the Water Industrythe U.K."
(London,Sept. 23-25, 1991) and on "Water Supplyand SewerageServicesin DevelopingCountries: from Technical
Assistanceto ConcessionContracts,"Paris, Sept. 26-27, 1991.

road infrastructure.In France, toll concessionsmade possible the creation of a 5500-km motorway
system (as of 1990). At its outset the system was operated with active support from the French
governmentin the form of contributionsin kind, financialaid and loan guarantees. Of the four private
companiesoriginallyinvolvedas concessionaires,three were later integratedinto the public sector in the
early 1980sas a result of financialproblems.Concessionairescurrently includeboth public and private
sector companies, and are chosen through a bidding process (Berthier, 1991). In Argentina, the
governmenthas in the last couple of years opened the highwaysector to extensive concessioning,and
consequentlyas of early 1992some 40 percent of the lengthof highwaysis being operated as toll roads.

35
As in the other types of contractsdiscussedhere, the advantagesor disadvantagesof such schemesdepend
on the incentivesimbeddedin the contractdesign. In Mexico,for example,recent contractsfor toll roads
give the private parties assurancesof traffic volumeswhich now appear overoptimisticand are creating
a burden for the governmnent
as guarantor.
Private entrepreneurship
New entry. Private ownershipof infrastructurecan come about, first, from entry by new
entities, which can be promotedby the removalof statutoryrestrictionson competition.In trucking, for
example,a numberof countries(e.g., Mexico,Hungary, Poland)have eliminatedthe former regulations
which imposed monopoly or cartel structures. As a result of this deregulation in Hungary,
privately-owned trucking companies supplied about 18 percent of national ton-kms in 1991 (from
virtually none in 1981) (Bennathan,Gutman, and Thompson, 1991). As discussed in Part One, for
activitieswith low sunk capital costs, deregulationmay be most appropriateto promoteprivatizationby
r.ew entry. Proactive measures may also be needed to facilitatethe growth of such new companies in
many developingcountries and enable them to challengethe dominantsuppliers in a given market (see
more on this in Part Three).
Joint ventures. Shared ownership and control of infrastructurethrough minority private
participationcan be a means of introducingexternalcapitaland know-how. As only one such example,
Air France and the Czecho-Slovakairline, CSA, signed a partnership in early 1992 giving the foreign
group headedby Air France a 40 percent share in CSA. Air France is providingassets in kind as well
as technical expertise, and the deal will give both airlines new access to routes and markets. Suchjoint
ventures have potential to develop private sector participationin infrastructureentities where even a
minority interestby the foreignprivate partner is attractivebecauseof particularcommercialadvantages
to be expected,such as accessto other central/easternEuropeanmarkets. Minoritystakes will only be
an interestingpropositionwhen the public enterprise is basicallysound, and where the foreign partner
can have sufficientconfidencein the government'sbehaviortowardsthe enterprise.
Another type of joint venture is seen in railways, where a private or mixed company
contributescapitalfor new investment,with part of the returns obtainedthroughthe right to developreal
estate owned by the railway. Examplesof this arrangement(sometimescalled development gain) are
found mainly in developedcountries and Hong Kong, and have been under considerationin Thailand.
There are also cases of joint venturesby railwayand telecommunicationscompaniesfor combineduse
of the rail right of way for the layingof cables; the railwayprovidesthe land and enjoys a returnthrough
some share in the telecommunicationsrevenues. In urban water supply/seweragenetworks and road
networks,there can also be benefitsfrom mixed use of the land rights of way for the layingof pipes and
maintenanceworks.
Divestiture.The transfer of ownershipshares in public enterprisesto private interests has
become an increasingpractice in manycountries.The one majorcase of total private ownershipof water
supply systems (except for enclave systems, i.e. those that serve a limited residential or industrial
complex)is that in England and Wales (see Box 6). The most rigorous attemptto evaluatethe results of
this trend is a detailed study of twelve cases undertaken recently by the World Bank (Galal et. al,
forthcoming).The study estirmatedthe net changesin welfareto severalparties (the government[former
owners], new owners, employees,consumers, and externalsector) attributableto three privatizationsin
each of four countries-United Kingdom,Malaysia,Mexico, and Chile. The case studies includedthree
privatizationsof telecommunicationscompanies,four of airlines, two electricitycompanies,a port and

36
a trucking company (the only non-infrastructure case was that of a national lottery). The research
concludedthat the net overall welfare changewas positivein each case except one of a Mexicanairline.
In drawing conclusionsfrom this research for the design of policy on privatizationof
infrastructuregenerally,the followingcaveats shouldbe kept in mind:
(i)

The privatizationof a given companywhich has a particular combinationof natural

monopolyand non-natural monopolyactivitiesis not examined as a different case
from one where the company is, for example, composedof entirely non-natural
monopolyactivities.Another way of stating this is that the regulatory requirements
of the privatizedcompanywere not examined.Therefore it is difficultto generalize
aboutthe results of privatizationof other enterprisesin the same sector or to other
sectors, for whom regulatoryconditionsmay be different;

(ii)

the countrieschosen for the case studies are high or middle-income.It is not clear
that the results wouldbe equallyapplicablein countrieswith less capacityto regulate
and less developedprivate markets.

The last major category of institutional options to be considered, self-help or

cooperative-typearrangements,is discussedvery briefly in the next section under "Participation".
B. Comparisonand Evaluationof AlternativeInstitutional Arrangements
This Sectionsummarizesthe experienceswith institutionaloptionsdescribedby relatingthem
to the five objectives,or benefits, soughtfrom involvingthe private sector.

Skilledmanagement. The opportunityfor dynamicand motivatedmanagers is improved

by movingfrom a governmentdepartmentto a corporatizedparastatal; the quality of managementmay
be further improvedwith an externalmanagementcontract.Sustainabilityof good managementremains
a problem in all of these options, however, either because of the government'stendencyto reassert
influenceon managers,the lackof continuityof managementunder contracts, or lack of training of local
managers by private contractors. Longer-term arrangementsfor private sector participationwith full
autonomy of operational decision-making(e.g. leases, concessions)and with specific incentivesfor
training therefore have greater potentialfor achievingthe benefit of improvedmanagement.
Productive efficiency. The commercializationof parastatals may create the minimum
conditionsfor improvedproductiveefficiencyby introducingan awarenessof cost accounting.The use
of service contracts, managementcontracts and leasingmay further encouragecost reductionswhich are
within the reach of managerswithout requiring significantnew investment(since investmentdecisions
remain outsidethe contractor's controlin these cases). Improvementsin efficiencyfrom the streamlining
of procedures,introductionof managementinformationsystems, etc., maybe more likely to occur in the
cases of management/leasecontracts with external parties who can offer new ideas and experiences.
However,these gains can also be underminedby continuedgovernmentintervention,especiallyin labor
matters, and from problems imposedby inappropriategovernmentactions regardingtariff policies and
investments(bothof whichalso remainoutsidethe controlof theprivate contractor).Again, arrangements
such as concessionswhich give one party responsibilityfor investmentsas well as operations help to

37
address part of this problem. Ultimately, productive efficiencydepends on the incentivescreated by
owners and the regulatoryregime.
Dynamic efficiency.Insofar as this dependson investmentin new technology,meetingthis
objective requires institutionalforms where decisionson capital investmentare made on commercial
rather than political grounds.This conditionis more likely to be met where responsibilityfor operations
and investmentare linked in one entity which faces commercialrisk, such as a concessionor private
entrepreneur.Again, the regulatoryframeworkcan affectthe extentto whichthe entity enjoysthe benefits
from technologicalimprovements.
Accountability to customers. Improvementsin service quality may follow from changes
in managementpracticesand from new investment,but expliciteffortsby entitiesto respondto demands
for servicedependon the incentivesto do so. Sustainedresponsivenessand accountabilityto customers
requires either significantpressures on suppliersfrom a competitivemarket or, in the absenceof this,
other opportunitiesfor clientrepresentationin decision-making.The abovediscussionsuggeststhat such
accountabilityis at least likely to be greater with those arrangementsin which suppliers face commercial
risk in relying on revenuesfrom customers-such as under lease and concessioncontracts, or private
entrepreneurships.
Flnancial autonomy and resource additionality. Transferringinfrastructureactivitiesfrom
a governmentdepartmentto a commercializedand corporatizedparastatal formallyseparatesthem from
state budgets. However, experiencein many countriesshows that evenunder this arrangement,claims
on governmentfor deficit financing,or attemptsby governmentto siphon resources from the parastatal,
oftenpersist. In comparison,leasesand concessionsare attractiveas they effectivelyfree the publicsector
from working capital financingand commercialrisk. These two institutionalforms do not necessarily
attract new sources of revenue to the sector, however, since contractors are compensatedfrom user
charges/tariffswhich couldhave been recovered directlyby the public sector-albeit political obstacles
oftenpreventthis from happening.To the extent that contractsfor private O&Mrelievegovernmentfrom
current financiallosses, they would also free public budgetaryresources for new investment.BOTsand
divestiture do involve an infusion of private (domestic or foreign) financing for new investment.
However, the net financialcontributionfrom the privatepartner dependson the particularterms of each
agreement, and often governmentsmust make substantialfinancialcommitments,on-or off-budget, to
settle these deals.
To conclude, the limitedforms of private sector participationcan have important,if often
transitorybenefits. Moreover, these formnsare attractiveas interim stages towards the developmentof
more comprehensivecontractingof operations,through leasingor concessions.In principle, all of the
objectivesof private sector participationcan be met with concessionsas well as with private ownership
and control. However, to determinethe conditionsfor successfulimplementationof these arrangements
in a givensectorand country, it is necessaryto considerissuesconcerningthe policyenvironmentrelated
especiallyto competition,participation,regulation,pricing,and financing.Theseissues-which havebeen
the root of problems with many attemptsat institutionalreform (see Box 7)-are discussed in the next
section.

rA

0..

~.

...

.....

...

~ ~ ~ ~ ~ ~ ~~9
~
~
..w
~ ~~~~~~~~~~~~~~~~~~~~~~~---

..

r.
~~~~~~~~~~~I8~~~c
~
~~~~~~~~~~~~~~,14

AA

00F

cr..
S.g.ii

~ ~~~P

~~~~~c
a~~~~~~~~~~~~1~~~~~a

-4

0t

cl~~~~
cr
O~~~~~~~~~~
~~~~~

9,

to

eq

..

C3

C2

, W,

W~4

a~~~~~~~~

39

III. Conditionsfor Ensuringthe Effectivenessof InstitutionalArrangements

Improvedthe performanceof infrastructurerequiresnot merelygovernmentwithdrawalfrom
direct delivery of servicesin favor of the private sector. More fundamentally,the role of government
needs to change. In all the institutionalalternativesinvolvingthe private sector, governmentretains at
least some degree of responsibilityfor sectoralplanning, policymaking, regulation,and financing, even
when ownershipis transferred. For example, in contractingfor services of solid waste collection, the
municipalityissuing the bids needs to have the competenceand authorityto develop, negotiate,manage,
monitor, and enforce contracts.In developingtoll road concessionsor BOTsfor waste treatmentplants,
for example, the government must carry out initial studies of each potential scheme and investigate
alternativeoptions that may be more attractiveeconomically.It must examine the financialviability of
proposedconcessions(and on what terms, i.e., need for governmentguarantees);determinethe detailed
specifications of tender documents; and check ultimate compliance. In many countries, the basic
transformationof government'srole is madeeven more complexas it is occurringin the midstof a major
realignmentof responsibilitiesamonglevels of government-national, regional/provincial,and local.
A. CreatingCompetition
In most infrastructuresubsectors,someactivitieshaveeither high sunkcosts (e.g., supplying
water supply mains) or services that are most efficientlyprovidedas a natural monopoly(e.g., airline
service between certain city pairs).' In these cases, what can be open to competitionis the right to
operate the monopoly-also knownas "competitionfor the market." The truckingindustrycomes closest
to lacking any structuralbarriers to competition,or being virtuallycontestable.Once the contestability
of given activitiesin principlehas been determinedfrom sector analysis,the role for policy is to ensure
that there are no practicalbarriers to competition.
Exploiting contestability. In addition to removing statutory restrictions against entry,
measuresare often neededto removeconstraintsthat limit the abilityof new, small operatorsto compete
on fair terms with incumbentsuppliers-particularlyregardingthe availabilityof credit, accessto foreign
exchange,tax treatmentof dividendsand profits, importduties, and general administrativerequirements.
In Central/EasternEuropean countries, for example, bilateral permits for access of trucks to the EC
represent a source of market power and should be auctionedto allow smaller new trucking companies
to compete. Governmentscould also facilitateentry in such activitiesby encouragingleasing companies
to be formedor other measuresto be undertakenthat promote a domesticmarket for reselling capital
equipmentand vehicles. In several Africancountries(includingBurkinaFaso, Burundi, Central African
Republic, Cameroon, and Tanzania, amongothers), the constructionequipmentfleets of the Ministries
of Works are being transformed into commerciallyoriented leasing firms. It is also critical that all
potentialbarriers to contestabilityat the specific"entry points" to the marketbe correctlyidentified.For
example,in deregulatingurban bus transit in the UnitedKingdom, allowingfree access of all operators
to bus terminalswas found to be a key factor in promotingeffectivecompetition.In the United States'
deregulation of the airline industry, access to sunk airport facilities was found to be the element
constrainingentry into otherwisecontestableroutes.

2 In certain cases, even if a single provider would be the most efficie.at solution, when the monopolist offers particularly
unsatisfactory service it may be preferable to forego the economies of scale in favor of multiple providers who would offer better
service, albeit at higher cost.

40

Sometimescompetitivepressure can be exerted against monopolysuppliers by developing

alternative services, even if these are not obvious substitutes. For example, water vendors, radio
communications,and freightlightersare oftennot consideredpart of theformal supplystructurealongside
the publicwaterutility, local exchangenetwork, or port facility, respectively.But they maybe considered
as viable service alternativesfor at least some users.
Competitionin Contracting.Contractingpracticesfor public works constructionin many
countriesare notorious for corruption (or in economicjargon, "rent-seeking"). In extending contracts
to cover operations and maintenance,it is especiallyimportantto ensure effectivecompetitionin order
to curb the potential for corruption. It is critical that the bidding process be transparent and that
there be a clear and formal separation between the authority issuing bids and potential contractors.
However,corruptionin somecountriesmaybe virtuallyinevitabledespitesuch controls.Ultimately,what
mattersis whetherthe servicesrequired arebeingdelivered,withoutan unsupportabledrain on the public
purse.
To give contractorsadequateincentivesto make improvementswith longer-term payoffs,
it is importantto permit contractsto be renewed, and to make the contractperiods long enoughto cover
the amortizationof invested capital. However, with these options, incumbentsinvariably develop an
advantagein bidding for subsequentcontracts. This potential cost may be worth taking when weighed
againstthe gains from a contractingarrangementwhich has a track recordof good service performance.
Some competitivepressure can be fostered at the outset by developing multiplepotential
challengers.For example,competitionbetweenthepublicsectorand private contractorsin municipalsolid
waste collectionhas resulted in substantialimprovementsin efficiencyof public crews. In Bangkokand
Bogota, for example, the market is divided into zones, with the private sector competingfor contracts
in certain zones and public agencies retaining the remaining zones. This has made both groups of
providers more accountablefor the cost and quality of their services; retainingthe minimalcapacity in
the public sector also protectscities from becominghostageto private cartels. It is importantto note that
in these cases, the various crews do not compete within the same area, but rather face competitive
pressure on their right to the contract.
The same principle operates in contractingfor municipalwater supply services in France
(describedin Box6). Althoughthe incumbentconcessionairesin France often win contractrenewalsover
challengers,the municipalitiesretain the option of taking over the operationby force accountif they are
not satisfiedwith the contractor's performance.This createsa margin of competitivepressure. The City
of Paris contracts with two private-sector water companies in different parts of the city, which also
provides a measure of comparative ("yardstick") competition. This proxy form of competition still
requiressome mechanismto createpressure from consumersto achievesatisfactoryservice. In France,
this pressure is exerted in large part throughvoters' active involvementin local elections-an optionthat
will eventually become known in the many developing countries where political decentralizationis
underway.
B. DevelopingEffective Participation
The needto make infrastructureservicesresponsiveto users poses a particularchallengein
the case of activities for which competitivemarkets are highly imperfect. For these activities, special
measuresare neededto ensure that all interestedparties can participatein the process of identifyingand

41

satisfyinginfrastructureservicerequirements.For muchof infrastructure,the relevantparticipantsshould

include not only users but other groups affected indirectly.
The potentialfor developingparticipationvaries with the nature of the service involved,the
scale and technologicalcomplexityof production, and sociopoliticalcharacteristicsof the population
affected. Encouraging participation is especially important for activities in which users have few
substitutesavailableOowpotential for "exit") and lack the means of expressingpersonalpreferencesto
suppliers (exercising"voice").'
One model of participationwith a considerablerecord of experience,particularly in rural
power and telecommunications,is that of cooperatives, wherebyusers act as the full or partial owners,
financiers, and suppliers. The history of rural cooperativesin organizingthe expansionof services in
some developedcountries (e.g., United States, Canada, and Finland for telecommunications)is being
viewed with considerableinterestby countriessuch as Poland. Successfulservicedevelopmentfollowing
these examples, however, requires certain conditions. In particular, a strong regulatory influence is
needed to ensure compatible technical standards for interconnection to the broader power and
telecommunicationsnetworks, and to ensure that monopolyproviders withingeographicalareas do not
overbuild redundantsystemsor abuse their position.
A second approach to broadeningparticipationis to incorporatethe views of clients into
decision-making,such as by involvingcustomerspokesmenamong utilities' boards of directors, open
hearings on regulatory reviews and proposed investments,and processes to provide considerationof
consumers' complaints.Broadeningpublic accessto informationabout infrastructureactivities,such as
throughpublicationof financialaccountsand servicerecords, is an importantelementof this approach.
The national roads authority of Sierra Leone, for example, includesroad users (e.g., the Chamber of
Commerceand the automobileassociation)on its governingboard. Althoughthe authority has only been
existing in its present form sinceearly 1992, the users' representativeshave proven influentialin making
importantpersonneldecisions.The Sierra Leone exampleis being consideredas a model for some other
African countries.
There are also many cases throughoutthe world where servicesof a largely public nature,
such as rural roads, communalwatersupply and sanitation,irrigation,and social infrastructure,are being
provided successfully by community-based ("self-help") organizations or by nongovernmental
organizations(NGOs). Indeed, in urban squatter settlements,self-help through the informalsubsistence
sector is the typical approachto provisionof shelter and basicinfrastructure.Typically,beneficiariesare
limited to a distinct geographicregion (e.g., urban neighborhood,village)and use simple technologies.
A samplingof some successfulparticipatoryapproachesto provisionof infrastructureis providedin Box
8.
Early efforts to generatebeneficiaryparticipationin rural developmentprojectshad mixed
results. More recent experiencesin the irrigationsubsectorwith water users associations(WUAs)have
been more positive, althoughthey are not universallysuccessful.Positive experiencesin workingwith
local NGOs in pilot rural roads projects in Ghana have helped to define a model approach to this
subsector.

X These concepts arc developed by Samuel Paul in Accountability in Public Services: Exit, Voice and Capture, World Bank,

Working Papcr No. 614, 1991.

42
.. ........
:. -... .......
......

. .. . . . . . . .

- ox 8Some ExampleIof Community Participan

.. .

. .... ...........................................................

*uthe Proviso of Iafrastructure

:linda (Brazil):In 1983, the muricipalgovernmentof OWa :(a city<f 400,000 in the metropolitanarei
of Recife)beganaproject for environmentalupgrading,beginningwitha pilot projectin a slum neighborhood.The local:
: communityi
assiti on hired local labor to carry out drainagcworks, usingqlow-costdrains and land fill; solid wastce
collection,includig compoting and recyclingactivities;sanitition, through constructionof ventilated imnproved
pit
latrines; and rad. upgradingwith low cost soil/cementpaving. A1l materials were produced through newly created:
micro-enterprisesian the local area.

-Neigbborhood
i
residentswere responsiblefor maintena:ceof the works.While initialcommunityconcerns
a communityassociationfor the larger area and throughpolical leaders,the Acities of the project
tisinulated the aion of a dwellers' council withmithe pilot slum neighborhood.The city supported community7..
i volvementAby
oreatinga specialprojectunit withinthe municipalurbanizationcompanythat providedtining
1 and public;
....
educato in.additionto urban uprding....
. ..
. .. .
.
.
.
i::j::were
voicedthrugh

e d

*-00A i -.:0-:T
Ne
h0 sustainabiliyof environmental management efforts in the pilot area is atributed to: the procedure
ofdireat contractingof works by the community;the developmentof communityawarenessabout the need to maintain
:::environmental
infrastructure; andthe mobilization of household resources for local investmnent.......

*.

0200it;itit
020Iharavi
.D00-iA(India). In 1979, this informal settlement of 400,000 people in Bombay formed a local
organizationcalled PROUDto improveits living conditions.Today, PROUDis a wcll-established,active organization:
made upof 150 neighborhoodcommitteesthat meet monthlyto discussloca problemsand plan solutions. Problemsthat
canot be solvedwithin thii committeeare referredto one of fivc area councils;thee icouncilare
i
ther represented 0
-nan ezocutivecommitteewhich fonulates policiesand implementsprogram.sinthe setulmcnt. An.annual convention:
open to all residentsreviewsprogresson issuessuchas water supply,drainage,solid wastecollectionand health.Because
of PROUD';strong base of support among residents, it has been effctive in pitioningk municipalauthorities for
en0>Ivinmental
improvementsand blocking plans for demolitions and developmentthat. would adversely affect the
community.

.. ......

::: Sulabh ShauchalayaProiect (India):This projectwas started in Biharin 1974as a vol

non-profit
communitiesmobilizedto improve their
organizations
rtAdby governent sponsorsh and aid, which responds1to:
sanitary conditions.The aim of the project is to convertdry householdltrines into new pourflush waterseallatrines,
and to constructpublictoilets and wash areas with the same technology,The Sulabhprojeoctapprovesgrant and loans to
It also eonstruts materialsin Sulabhworkshops educates
homeownersfor privatelatrines,and administersthe fmnancing.
thecommunityaboutpublichygiene,cleansthe latrines for a servicefee, and retrainsthe scavengers(traditionalcleaners
of the former dry latrines) in maintenance of the new latrines or in other worL,.

.. .

..

Oranei Pilot Proiect (Pakistan) The OPP was startcd ithhcore external fundingto develop affordable
sewerage:for Karachi's squatter settlementsand create organizationsto providethetsysterns. Technicalinnovationsin
design, coupled witheliminationof corruptionand provisionof communitylabor, led to significantcost reductionsfor
in-house sanitarylatrines and sewers on houseplots, and undergroundsewers on the streets. The projectstaff provides
public educationand technical assistanceto residents,whose responsibilitiessrc to finance their share of the costs,
especially in constructionand elect 'lane managers". The latter elect neighborhoodcommitteeswhich manage the
secondarysewers. OPP organizationshave been able to pressurethe municipalityto provide funds for the construction
of primary and secondarysewers.The OPP has led to provisionof seweragefor over 600,000poor residentsof Karachi
and the approachis being followedin other municipalitiesof the country.
(Excerptedfrom: 'EtnvironmentalStrategiesfor Cities",UNDPIWorldBanktUNCHS(Habitat),October10, 1992draft).

A review of the experiencewith participatoryprojects supported by internationaldonors

43
suggestsomekey lessons.' First, an importantdistinctionmust be made betweenparticipatorystrategies
that are limitedto consultationwith interestgroups, such as through demandsurveys, as contrastedwith
strategiesthat offer significantinvolvementin decision making. Experienceindicatesthat consultation
can help in identifyingfeatures of a project that will contributeto its effectiveness. However, more
substantiveparticipationby users is associatedwith greater and more sustainablebenefits, whetherthese
are defined in terms of efficiency and cost recovery or, especially, in terms of longer-term
capacity-buildingor empowermentof the communityinvolved.A furtherdistinctionis necessarybetween
consultationor involvementof groups only at limitedpointsin the projectcycle, rather than throughout
all stages of the activity and particularly during the initial identificationand design. There are
numerous examples of poor performance in projects where attempts were made to involve users in
operation, maintenance, and cost recovery, but withoutbringing them into early decisions regarding
project design and location.
Participationoccurs more strongly in practice when it is an explicitobjectiveof the project
and staff are rewardedfor its achievement.Professionallytrainedand paid staff are foundto be important
to ensure sustainabilityof participatoryassociations,such as WUAs. In addition, developmentactivities
that are open-ended, flexible,adaptiveprocessesof respondingto the evolvingconcernsof participants,
where programs reflect learning from experience and pilot demonstration, are more conducive to
maintainingparticipationthan are projects with defined timetables and predetermined"blueprints" for
implementation.To involve women and the very poor members of the community,projects must be
designedat the outset to recognizetheir different concernsand capacities.Accordingto a recent research
project on 122 completedrural water supplyprojects, the degree of participationby women, and a high
level of client participation generally, were found to be highly correlated with overall project
effectiveness.?'

C. ReducingRisk
An obstacle often mentioned to involving the private sector in infrastructure in poor
countriesis that private entitiesdo not have the means to enter the market, and that neither the domestic
nor foreignprivate sectorhas sufficientinterestin assumingthe risks (commercialand political) involved.
Risks need to be balancedby potentialrewardsand by autonomyof decision-making.Private firms will
want to be assured that revenues are adequateto cover costs and allow for a reasonable profit. The
commercial risk can be reduced by clear policies on tariff-setting and revenue allocation, and by
disclosureof informationon the conditionof assets and effectivedemand.Both commercialand political
risks are more acceptablein a stable macroeconomicenvironmentthat minimizesuncertainty. It is also
preferable for the rights and responsibilitiesof governmentand private partners to be clearly defined;
sometimesthis maybe easierwith legal transformationof sector institutions,as in the cases of CIElEECI
in Cote d'Ivoire and SONEG/SEEGin Guinea (see Box 5).
In the case of activitieswith large sunk capitalexpenditures,private investors are unlikely
to takeownershipof such assetsin most countrieswithoutbeingassuredof monopolyprofits, a condition

n See C.O.N. Moser, 'Community Participation in Urban Projects in the Third World", in Progress in Planning, Vol. 32, p.
71-133, 1989; and Deepa Narayan, 'The Challenge of Participatory Development: Lessons Learned in Rural Water Supply",
UNDP-World Bank Water Supply and Sanitation Program, March 1993.
23 Narayan, op. cit..

44

that is often unacceptableto government and consumers.' Similarly, BOT-type arrangementsin a

numberof cases have involvedgovernmentguaranteesof revenuesor "take or pay" agreements,which
insulate the private investor from commercial risk and thus create unfavorable incentives. Private
ownership of long-lived assets may not be a desirable or practical alternativein many countries. The
more narrowly-focusedoptions for private operation (service contracts, management contracts, and
leases) may be more practical alternativesto gain many of the benefits of private-sector involvement.
Moving progressivelyfrom these types of contracts to a full-scale concessionarrangement, or even
eventuallyto divestiture,may be a good strategyduring a transitionperiod, as it builds naturallyon the
private operators' experiencewith the market and allowfor gradual expansionof their role.
To develop a viable private contracting industry, the public sector must be capable of
sustainingdemandfor the contractors' services.This requiresadequatesourcesof financing,either based
on tariff revenues,budgetarytransfers (for nonremunerativeactivitieswith a socialjustification, such as
public service obligationsfor urban bus operations),or foreign aid on a decliningbasis to support initial
stagesof private-sectorinvolvement.A particularlydifficultarea of risk arises in countrieswhereprivate
operators are unable to enforce payment of bills by public-sector entities. In such cases, private
contractingmay not be feasible until broaderreform of public enterprisesenforcescommercialbehavior
throughoutthe sector.
Consumersand the government,as the owner of assets, must also minimizethe risks that
contractors will provide inadequateservice or fail to maintain the assets during the p'-riod of the
agreement. These risks may be mitigated by strict monitoring of performance based on indicators
specified in the contract, adequatetariff adjustments,and mechanismsfor feedback from consumers.
Many countriesdo not have the capacityto ensure these conditionseffectivelyat the present time. Here
again, a realistic weighingof the risks is needed: in such cases the quality of service and maintenance
record of the existingprovider (usuallypublic monopolist)are often so poor that, in comparison, such
risks may be acceptableto the public.
Where divestitureis a practical option, measuresmay need to be taken by governmentto
prepare an entity for private ownershipby removingundue risks, such as by supportingthe reductionof
redundantlabor. Overstaffingis a nearly universal phenomenonin public infrastructureagencies and
enterprises, particularly in transport, where job creation is a traditional source of political favors.'
Research on strategiesof reducing excess labor in transport entities in several countries has found that
when dismissals are combined with severance pay, the reduction in force is more likely to be both
politicallyacceptedand sustainable.In cases of privatization,governmentfinancingof at least part of the
severance pay may be necessary,but the actual reductionsin force should be implementedby the new
owners, who have the best knowledgeof their labor requirements.A severance program with partial
government finding was a facilitating factor in the massive labor reductions involved in the recent
privatizationof Argentina's railways.

29There may be cases where monopoly rents are worth accepting: for example, where there is a high unmet demand for modem
telecommunications, certain users such as producers of high-value exports may be willing to pay a monopolist's premium for
services rather than go without them entirely.
30 Alice Galenson, -Labor Redundancy in the Transport Sector," World Bank INU Working Paper No. WPS 158, 1989; Jan
Svejnar and Katherine Terrell, 'ReducingLabor Redundancy in State-Owned Enterprises", INU WPS 792, 1991.

45

D. Effective Regulation-"Doing Well Only What is Necessary"

The worldwide experience and thinking in this area has been rapidly evolving in recent
years. Models of regulationthat have been developedfor the particular institutionalcircumstancesof
certain countriesmay need further adaptationin other countrieswhich are undergoingwidespreadand
rapid sectoral change in the context of generallyweak institutions.This paper argues that it is essential
to first define explicitly the characteristicsof any infrastructureactivity that makes a case for public
interventionthrough regulation,and then to identify the objectivesto be served. In many cases, such as
where there are no structuralbarriers to competition,the appropriaterole for governmentis to liberalize

'-Box9 Cycles of Public Sector and Private Sector Involvement in Urban Bus Servce
Many cities in both developedand devclopingcountries have cxperienceda cycle of privite and public
involvernentin urbanbus service.The cycle, which can be dividedinto ten phases, beginswith the private sectorbeing
largelyresponsiblefor urban bus service, followedby an increasingdegree of public involvementand, at least in some
'cases by a patal or completereturnto privateprovision.
The fst phase is the entrepreneurialstage when the industry first emerps, At ihis stage srvie i
typically provided by numerou suall firms, often with only a few vehicles in their flet sh
second Dhac is
rbaracterizd by mergger and consolidationof small firms into a few dominantcompanie, often with lttle overlap
betweentheir route networks.In sone cities,the emergenceof tailway(tram)technologywas partly responsibk fr t
consolidation.
Consoldationis typicallyfollowedin phasethreeby regulationof faresand fanchising of firmstocontrol
routes and ertry. Thes were often a responseto perceivedmarketpower of the firms created in phaie two to publi
outcry against farc increases; andlor to eliminateperceived -chaoticwcompetition.'Te fourth phase involvesa graduat
downwardtrend in thc proftability of tho pivate firms, followedby i fifth phu of capital withdrawaland service
rduedoin. Sometirnesthis phasehas reflectedthe reluctanceof regulatorsto pemit adequatefare incre8ses,adjustment
of routes, and ervices needed to maintainpro'fitability.The available
become poorly m ain and: service
deteriorates.

neu

s'ubli authoritis t eprver the

t,
'r vatebus
us'b a X
fol wd
public subsidiesis madeto tesore capital and services.The eihth Dhase
consists of incr alnit
operatingefficienciesand real unit costs, often due to overstaffingand poor operting practices.The subsidy
burde6nget
worse, until in the ninth phasc, public auithoritiesface the choice betweensignificantfare increaes tocurb t subsidy
drain, or service eductions,

the

in which an inusion

of

-n sm
onities, this diemma is resolvedby enteringa enthhase, in which the bus:servs, reoturne
us
to the private sector. In most cases, public regulationof fares is retained, resumingthe cycle in its fhird phase.
Source:John R. Meyer and Jose A. Gomez-lbanez. TransitkBus Privatizationand Deregulationaroundthe World: Some
.Perspectivesand Lcssons," in Jose Carbajo, cd. (1992).

entry and to ensure commercial freedom in the market rather than regulate specific behavior.
Alternatively, where there are structural barriers to entry or conditions of natural monopoly, some
regulationis neededto ensure reasonableterms for accessto network infrastructure,and to guard against
abusesof monopolypower.Otherproblemsthat haveprompteda regulatoryresponsein many cases, such
as high costs and poor service, often persist and worsenunder regulationas practiced(see Box 9 on the

46
urban bus sector).Thisdiscussionrefers mainlyto "economicregulation"-that pertainingto the structure
or behaviorof enterprises,e.g. regardingentry/exit and pricing-although manyof the points here also
apply to administrativeregulationfor safety, health, or environmentalreasons.
3 ' A minimumrequirementfor effectiveregulation
The Legal Framework.
of infrastructure
is a frameworkof law pertainingto propertyrights, liability, conflictresolution, and contracting.There
must also be capacity to enforce the laws, and credible assurancesthat they will not be changed by
political whim. Many countrieshave had legal or constitutionalbarriers to the transfer of rights to own
or operate water resources or basic telecommunications.Others have legal limits on the government's
ability to make financial commitmentsbeyond a given fiscal year or budgetary allocation, which can
constrainthe use of multi-year contracts.

Determiningreallatory functions anid authority. Most experiences to date with formal

regulationsuggest the followingguidelinesin structuringsuch regulation.
(i)

Responsibilitiesfor regulationand for operation should be formally separated. This

distinctionis also importantin the effective enforcementof environmental,health,
and safety regulations, which can impose financial costs on the operators. In the
United Statesand UnitedKingdom, the regulatoryauthority is commonlya legally
independententity; in French urban public services, the necessary distinction is
enforcedby the municipalities(regulators)delegatingthe operationalresponsibilities
through contracts.

(ii)

The regulatory processes should be capable of straightforward and prompt

implementation-for example,rules for tariff review should be based on formulas
triggeringautomaticadjustments;the periods betweenreviewsshould not be so short
as to impede the enterprises' managerialautonomy;and the reporting requirements
shouldbe as limited and simple as possible.

(iii) The method of tariff regulationshould enable producers to enjoy the benefits of
efficiencyimprovements,and not requireregulatorsto have full informationon costs.
(iv)

Rulings should be enforceable,with recourseto appeal.

(v)

There should be opportunity for interested groups, including users of the

infrastructure,to present views and be informedof the decisions.Regulationis also
more effective when it coexists with other strong institutions such as private
insurancemarkets (which can support observanceof safety and health regulations)
and an independentjudiciary.

In countrieswhere the overall institutionalcapacityis weak, the above guidelines need to

be interpretedcarefully. A recent comparativestudy of the institutionalcontext of telecommunications
regulation in several developing countries and the U.K. suggests that one of the most important
considerationsin designingregulationunder weak institutionsis predictability.Regulatorymethodsthat

Further discussion of the following points is provided in a background note, 'The Legal Framework and Regulation of
Infrastructurem by Rita Hilton (INURD), draft, 1992.
31

47
allowvery precise criteria for decisionmakingare found to be more effectivein Chile, for example,than
the approaches which give regulators more discretion, such as those used in the United Kingdom.
Similarly, procedures for public hearings through a U.S.-style public utility commissionwere very
unsuccessfulin the political/institutionalcontextof Jamaica.There is need in such cases to find ways of
reflectingthe public's concernseffectivelywithoutunderminingthe private operators' confidencein the
stabilityand impartialityof the regulatorysystem.' In this connection,it is very importantfor regulators
to have informationon user satisfaction,such as indicatorsof service quality and consumersurvey data.
Regulationthrough contractualoversieht. In the case of lease and concessioncontracts, the
basic parameters for regulation (the formula for setting tariffs, the mechanism and frequency of
adjustment,and the output to be delivered, e.g. a given quality of service and conditionof the facilities
to be maintained)are specified in the contract as negotiated.It is often advisablefor the contractorto
collectfees directlyfrom users and turn over the agreedportionto the contractingauthority;this provides
strong incentivesfor efficientoperationto reducecosts, provides high service quality, and achievesgood
collectionrates. Contractor's fees needto be adjustedperiodicallyto reflect inflation,and it is important
that the adjustmentmechanismbe transparentand easy to administer.For lengthy contractperiods such
as concessions,the fee review may need to take accountof longer term changesin parameters such as
demand, nature of service, or input mix.
It has been argued earlier that contractingfor services may be a good first step in the
transitionto broader private-sectorinvolvementthroughleasing, concessions,and eventuallydivestiture
in some cases. Oversightof service contractingis not a negligibletask in itself. But for a government
agencyor parastatalunaccustomedto formalregulation,focusingon the practiceof individualcontracting
may provide a gradual learning process. Contractingprovides the public authority with significant
leveragefor enforcement,focusedon the performanceindicatorsspecifiedin the contractand conditions
for its renewal. Moreover, as the practice of contracting expands, the government not only gains
experiencewith regulatoryfunctions,but the private sector also gains an increasingstake in the process.
In brief, institutingregulationthrougha transitionalprocessof expandingthe practice of contractingmay
be a more practical approachin many developingcountriesthan startingfrom scratch with a full-blown
independentregulatory agency.
Costs of regulation. A final point to be considered is the financial costs of regulation,
including contractualoversight. In the case of public transport, studies indicate that about one-fifth of
the savings in operating costs achievedby shifting from public bus service to competitivetenders is
absorbed by necessary regulation (Scurfield, 1990). In the municipal solid waste (MSW) subsector,
studies in developed countriesindicatethat private collectioncan be 20-40 percent less costlythan the
public services, but the costs of monitoringcan averagebetween 10-25 percent of the service contract
costs (Cointreau-Levine,1992). In leases and concessions,the imputed costs of regulation should be
coveredby the share of revenueremitted by the contractorto the public authority.This approachwould
focus attentionon the costs and benefitsof regulation,forcingthem to be judged in light of overall costs
and tariffs in the sector.

32 Brian Levy and Pablo Spiller, 'Regulatory Institutions and the Performance of Private Telecommunications: A Comparative

Analysis of Five Country Studies", January 29, 1993 draft.

48
E. Pricing and Flnancing
The nature of pricing (i.e., tariffs or user charges) and incentives created by financing
schemeshave an importantaffect on the allocativeefficiencyof infrastructureinvestment,as well as on
the internalsustainabilityand productiveefficiencyof services.This section addressesissuesof pricing
and financingin those cases where there is a legitimaterole for public involvement.
Pricing. As an overallobjective,tariff revenuesshouldbe sufficientto coveroperatingcosts,
debt service, depreciation, and the administrativecosts of investmentplanning and regulation. Much
analysishas been devoted to the methodologicalissues in setting prices for infrastructureservices, and
this work is not reviewedhere. One key issue is that marginalcost is generallyrecognizedto be the most
efficient determinationof price; however, it does not always generate sufficient revenue to finance
services involvinghigh fixed costs of production. In many cases, however, the tariffs charged are not
sufficientto cover even variablecosts.
Various pricing tools may be needed to deal with the sometimes conflictingconcerns of
efficiency,cost recovery, and equityfor each typeof infrastructure.Thesealternativepricing approaches
take accountof specificdemandfactorsas well as costsof supply. The use of a two-part tariff structure,
for example, can help to meet both efficiencyand financingobjectives;this structure involvesa fixed
component coveringthe marginalcost of providingaccess to the service network, and a variable part
based on the volumeof consumptionof particularservices. Other approachesinclude so-called Ramsey
pricing, in whichhigherprices are chargedto users whosedemandis most price-inelastic.In theory, this
reduces the welfare loss of not chargingmarginalcosts, althoughit can have unfavorabledistributional
implications, since the poor have inelastic demand for some types of services (e.g., urban public
transport).
Rising block pricing (higherrates for larger volumesused) and congestionpricing (higher
rates for serviceswith higherdemand)are also potentiallyattractiveoptionsin some circumstances(e.g.,
to discouragewaste of water or to managetraffic on certaintransportroutes). They require considerable
informationabout demandto design and implement.For example,in Kumasi, Ghana, surveysof water
users revealedthat the large volume consumers, who would be subject to higher rates under a block
scheme, includedpoor householdssharing connections.3In manycountries, the servicesof utilities are
not metered, which limits the scope for using price to affect to user behavior.
An additionalissue concernsthe structure of tariffs across servicesfor a given sector. As
a principlefor policy, the rate structureshouldreflectlong-run incrementalcostsof each type of service.
It is typicalin manyinfrastructuresubsectorsfor tariffsto reflectinternalcross-subsidies,e.g. high tariffs
for long-distancetelephoneserviceto financeunderpricedlocal service;or high watertariffs to industrial
users to subsidizehouseholds.As the institutionalreforms discussedhere lead to facilitiesand services
becoming 'unbundled" (managedby different entities),and increasedentry from the private sector, the
structure of rates will have importantimplicationsfor the incentivesfor investmentand operation in
particular market segments.Social objectives,such as the concern to ensure the population access to
essential services, are preferablyfinancedby explicitbudgetarysubsidiesrather than by the untargeted
transfer paymentsoften impliedwithin existing tariffs structures. Before subsidiesfor particular user

33

See Whitington, Dale et al. (1992).

49

groups are considered,however, their "willingnessto pay" shouldbe well investigated-includingtheir

separatedemandfor connectionto servicesas opposedto demandby volumeof consumption.
There is a further issuewhen liberalizingentry enablesprivateoperatorsto "cream off" the
most profitable markets, as tends to occur with service contractingof MSW and urban transport, for
example. The need to provide essential services in poor neighborhoodsmay therefore require some
revenue sharing among operatorsin different markets, or taxationof the private providers in wealthier
areas. Such an arrangementwould have to be clearlyestablishedas a conditionof the service contract.
Similarly, countrieswhich must expandtelecommunicationscoverage from an extremelylow base are
constrainedif profits from the marketsalreadyserved are lost to further investmentin the sector. In such
cases, govermmentsshould ensure that new private entrants share some of the legitimatepublic service
obligationsin the sector, as well as responsibilityfor needed new investment.
There is a strong argument for financing infrastructureservices through taxes when the
administrativecosts of these revenuesare lower than for direct user charges.This is the case with public
goodssuch as street cleaningand traffic signalling,as the consumptionby individualscannotbe priced.
Tax financing is not appropriatewhen it leads to unfavorable incentivesfor resource allocation-as
occurs, for example, when all farmers are taxed on their output prices rather than charged specifically
for the use of irrigationwater. Whereverpossible, the incidenceof taxationshouldfall on the population
benefitting from the infrastructure(except where there is a case for reverse transfers, i.e., targeted
subsidies to certain users). Taxes on vehicles and fuel (a partial proxy for road user charges), and
sanitationchargeslinked to property,generallyfit these conditionsas forms of "benefittaxation".' The
explicit identificationof such revenues as user charges, however imperfect, is an importantelement in
transforminginfrastructureactivitiessuch as road maintenanceandmunicipalsanitationfrom bureaucratic
financing (dependenton general budgetarytransfers)to a commercialorientation.
Debt and Equity Financing. Financing capital expenditures for infrastructure through
borrowing is essentialto ease the cash flow problem of large investments;it also promotes an equitable
burden-sharingbetweengenerations,as the benefitsof long-lived assets accrue in the future. The major
obstacles to debt financing are the lack of creditworthinessof many public suppliers of infrastructure
(especially at the local level), and the immature financial markets for long-term capital in many
developingcountries.In manydevelopingcountries,expandingprivate involvementin infrastructurewill
require greater availabilityof even medium-termdomestic credit, particularly for the small firms that
could providemuch of the service contractingand new entry in subsectorssuch as trucking.
The specializedfinancialinstitutionsthat have been set up in manydevelopingcountriesto
provide credit to municipalitiesfor infrastructureprojects have a mixed record (Davey, 1988). The
lessonsof experiencewith these institutionssuggestthat they should be viewedonly as interimsolutions
to the problem of municipal capital finance; that is, they should help to develop a track record of
responsible borrowing by municipalities, while private long-term capital markets emerge. Where
municipal credit institutionsare establishedas a transitionalmeasure, the governmentshould let them
operate on solelyfinancialcriteria. At the same time, centralgovernmentsshouldhelp local governments

34A 'benefit tax" is a tax whose incidence is confined to the beneficiaries of the services it finances, but which is not linked
to specific consumption (and therefore is not a price). A property-related charge for street repair or solid waste collection in
a given residential area is an example of benefit taxation.

50

to become good credit risks, including by developing appropriatepolicies on the sharing of fiscal
revenuesand allocationof expenditureresponsibilities.
Financinginfrastructurethrough instrumentssuch as revenue bonds and equity issues can
provide a good "feedstock"for emerging capital markets and attract funds from institutionalinvestors
(both foreign and domestic). There is especially strong potential for these instruments in power,
telecommunications,and railways,oncethe sectorsdemonstratefinancialdiscipline.An appropriatelegal
and regulatoryframeworkas discussedaboveis essentialto foster capitalmarketactivityin infrastructure.
In addition, participationby multilateralinstitutionssuch as the World Bank and IFC can be important
to assure the privatesector that the environmentis sufficientlysoundfor long-term financialinvolvement
in the sector.
F. Investment Planning
The planning of infrastructuredevelopmentshould be based on analysisof the nature of
demand for specific services, not on quantitativeprojections of physical "need." The design of a
demand-based strategy for infrastructurepolicy imposes additional informationrequirements.It must
involveanalysisof the underlyingdeterminantsof demand, such as the compositionof user groups and
their demandfor specific kinds of services (which depend, for example, on price elasticities),and the
patterns of congestion. Moreover, since demand is ever-changing, it is necessary for infrastructure
decisionsto be based on a dynamic processof assessingdemand-such as through competitivemarkets
wherepossible,regular demandsurveys,and/orchannelsfor participation.Makinga thoroughassessment
of demand also requires that plannerstake account of all possible alternativesto generate the flow of
servicesdemanded-includingmeasures(withor withoutinvestment)to increasethe efficiencyof existing
facilitiesand relieve specific sources of congestion,and to promote conservation(reducingdemand).
To practice a demandorientationin planninginvestments,as well as in their operation and
regulation, it is also necessary to define performance so as to reflect quality of service and user
satisfaction.Most of the performance indicators customarilyused by planners and operators in the
infrastructuresectors reflect physical parameters of the facilities and internal (including financial)
efficiency. In addition to these, service quality needs to be measured and monitored as an input to
assessinghow well sectoral,operational,and policyobjectivesare beingmet; to evaluatethe effectiveness
of alternativeservice providers; and to help planners and regulators to evaluate success in achieving
operating rules and standards, or performance benchmarks. Contracting for services, for example,
requires quite different performance indicators to measure output than contracting for public works
construction.

51

ANNEXI

GLOSSARY
(Note:This glossary is not intendedas a definitivelexicon, but only to explainhow terms are used in the
present paper. Some of the terms here may be used differentlyby others.)
BOT Build-Operate-Transfer.A form of concessionusuallyreferringto totallynewprojects. Typically
in a BOT, a private party (or consortium)agrees to finance and constructa facility, and operate and
maintainit, for a specifiedperiod and then transferthe facilityto a governmentor other public authority.
Variations include BOOT (Build-Own-Operate-Transfer)and BOO (Build-Own-Operate);in the last
case, the contractaccords the right to constructthe facility, which is not transferredback to the public
sector.
CommercializationStatusof a Stateenterprisewhichis financedmainlyby internally-generatedrevenues
(tariffs)and thus has financialautonomyto operateas a business;its accessto governmentsupport is very
circumscribed,e.g., limited to explicitcompensationfor public services.
Concession Arrangementwherebya private party leasesassets from a public authorityfor an extended
period, and has responsibilityfor financingspecifiednew fixed investmentsduring the period; these new
assets then revert to the public sector at expirationof the contract.
ContestabilityThe practicalthreat of competitionfrom new entrants in a market. In activitieswith high
contestability,exit by a produceris relativelycostless. The key criterionfor contestabilityis the absence
of sunk capital costs incurred in the even of exit.
CorporatizationTransformationof an enterprise or agency into a legal entity subjectto companylaw,
with formal separationof ownershipand managementresponsibilities,e.g. through a board of directors
or other body.
ExcludabilityThe ability to control accessto a good/service. High excludabilitymeans it is relatively
easy (not costly)to prevent users from consumingit.
GovernmentdepartmentArrangementwherebyservicesare providedand executedby civil servantsand
accountsare integratedinto the governmentbudget.
Leasing (or "affermage")Arrangementwherebya privateparty (essee) contractswith a publicauthority
for the right to operate a facility(or the right to a flow of revenuesfrom providinga specificservice)for
a specified period of time; this right may be called a franchise or license. The facility continuesto be
owned by the public authority. Unlike in a concession,the lessee does not have responsibilityfor
investmentsin fixedassets.' Financialrisk for operationand maintenanceis borne entirelyby the lessee.
Management contract Private contractor assumes responsibility for full range of operation and
maintenancefunctions,with authorityto makeday-to-day managementdecisions.Compensationmay be
based partially on services rendered (as for service contracts), and partially on performance achieved
(such as profit-sharing), which implieslimited financialrisk.

1 A lease may sometimes be called a 'service concession", and a BOT sometimes called a 'public works concession".

52
NetworkexternalitiesPositiveeffects of consumptionby one user on other users. E.g., the decisionby
one user to link to the telephone network yields benefits to other users, by increasing the range of
possible communicationsthrough the telephonesystem.
Private entrepreneurshipOwnershipof assets held in majority or totality by private agents, either
through divestiture(transfer/saleof shares or assets from public ownership)or creation of new private
entity. It is assumedhere that control of assets is proportionalto ownership;however, it is possible in
cases of divestiturethat majoritycontrol(of votingrights) is grantedto privateagents evenwhenthe State
continuesto hold the majorityof shares. In such cases a "majority privately-controlled"entity wouldbe
considereda private entrepreneurshiphere.
Parastatal(also public or State enterprise) An organizationowned and controlled in majority by the
State.
Public utility A public enterprise/parastatalin the infrastructuresectors (usually referring to power,
telecommunications,water supply, or sanitation).
Public (or State) enterpriseSee Parastatal.
"Self-help"Includesvariety of arrangementsfor financingand managementof projectsby beneficiaries,
such as by cooperatives,communityorganizations,or user associations.
Servicecontract(or "contracting-out")Arrangementwith private sector to perform particularoperating
or maintenancefunctionsfor a fixed period and for specifiedcompensation(no financialrisk taken).
Subtractability(or "rivalry") The impactthat consumptionby incrementalusers has on consumption
opportunitiesof all users. Low subtractabilitymeans that consumptionby one user does not impede
availabilityby other users (suchgoods/servicesare "jointly consumed").High subtractabilitymeansthat
consumptionby one user imposesidentifiable,calculablecostson other users (such goods are "privately
consumed").

"Traditional"parastatal A public enterprise with limited managerial autonomy; accounts are kept
separate from the governmentbudget, but with wide access to deficit financing from the budget and
governmentloans.

ANNEX 11 - CHARACTERISTICS OF INFRASTRUCTURE SECTORS/SUBSECTORSA BREAKDOWN BY ACTIVITY

ANNEX 11Page I of 16

SECTOR/SUBSECTOR: AIRPORTS/AIRWAYS
ACTIVITY

NATURE OF GOOD/SERVICE

Degreeof

Degreeof

Subtracta-

Excluda-

bility

bility

PRODUCTION ASPECTS

Degree

Economies

Degreeof

of

of

Coordina-

Sunk-

Scale

tion

ness

Runways, gates

Air trafficcontrol

Groundservices

moderate, L

low)

USO

~~~~~~~~~~SOCIAL
OBJECTIVES
INVOLVED

Necessary

Airplane services

Definitions: (H = high, M

EXTERNALITIES
OR

National
defense
concerns,
Noise, USO
Publicsafety,
liability

universal service obligations

Subtractabilitv(or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all users. Low subtractability (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability = = consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability

= =

it is relatively easy (i.e. non-costly) to prevent users from consuming it.

Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concems,
etc.

ANNEX II Page 2 of 16
SECTOR/SUBSECTOR: ELECTRIC POWER
NATURE OF GOOD/SERVICE

ACTIVITY

PRODUCTION ASPECTS
r
Degree of
Economies
Degree
Coordination
of
of
Necessary
Scale
Sunkness

EXTERNALITIES
OR
~~~~~SOCIAL
OBJECTIVES
INVOLVED

Degreeof
Subtractibility

Degree of
Excludability

Transmission

Network
effects

Distribution

Some arguments

Generation - thermal

- hydro, nuclear

Definitions:(H = high, M = moderate,L

Pollution

for USO

low). USO = universal erviceobligations

opportunitiesof all users. Low subtractability(nonrivalry) =

Subtractabilitv(or rivalry) refers to the impact that consumptionby incrementalusers has on consumption
consumptionby one user imposesidentifiable,
consumptionby one user does not impedeavailabilityto other users (such goodsare jointly consumed). Highsubtractability
csaculablecosts on other users (such goods are privately consumed).
Excludabilityrefers to the feasibilityof controllingaccessto a good. High excludability=

it is relativelyeasy (i.e. non-costly)to preventusers from using it.

in the event of exit from the activity.In activitieswith high contestability,entry and exit are relativelycostless (key
Sunknessrefen to extent of sunk capitalcosts incurred
of sunkcosts incurredin the event of exit).
criterionfor contestabilityis the absence
Coordtion refers to need for controlover the rightsto performthe function,e.g., becauseof technicalconditionsfor efficiency,need for equipmentstandards,safetyconcerns,
etc.
Smu11for diesl gcnerators,largefor steam generators.

ANNEX11Page 3 of 16
SECTOR/SUBSECTOR: IRRIGATION (GROUNDWATER)
ACTIVITY

NATURE OF GOOD/SERVICE

PRODUCTION ASPECTS
______

Degreeof
Subtractability

Degreeof
Excludability

Degree
of
Sunkness

Economies
of
Scale

Degree of
Coordination
Necessary

t
Deep Tubewell

HI

ShallowTubewell

EXTERNALITI
ES
~~~~OR
SOCLAL
OBJECTIVES
INVOLVED

Groundwater
depletionand
pollution
Waterlogging
and
Salinization

Definitions:(H

high, M

moderate,L = low) USO = universalservice obligations

Subtrmctabilitv
(or rivalry) refers to the impact that consumptionby incrementalusen has on consumptionopportunitiesof all users. Low subtractability(nonrivalry)= =
consumptionby one user does not impedeavailabilityto other users (such goodsarejointly consumed). High subtractability= = consumptionby one user imposesidentifiable,
calculablecosts on other users (such goodsare privately consumed).
Excludabilityrefer to the feasibilityof controllingacocssto a good. High excludability=

it is relativelyeasy (i.e. non-costly)to preventusers from consumingit.

Sunknessrefer to extent of sunk capital costs incurred in the event of exit from the activity. In activitieswith high contestability,entry and exit are relativelycostless(key
crterion for contestabilityis the absenc of sunk costs incurredin the event of exit).
Coordinationrefers to need for controlover the rightsto performthe function,e.g., becauseoftechnicalconditionsfor efficiency,need for equipmentstandards,safetyconcerns,
etc.
Any distributionor terminalsystems associatedwith tubewellsexhibitthe ame characteristicsas those for surfacewater systems.
' The degree of subtractabilityassociatedwith a givenwell actuallydepcndsuponthe natureof the watershed/aquiferdrawnfrom. In this instance,relativelyhighresourcescarcity
is assumed.

ANNEXn Page 4 of 16
SECTOR/SUBSECTOR:IRRIGATION (SURFACE WATER)
ACTIVITY

NATURE OF GOOD/SERVICE
Degreeof
Subtractability

Degreeof
Excludability

PRODUCTION ASPECTS
Degree
of
Sunkness

Economies
of
Scale

Degree of
Coordination
Necessary

OR
OBJECTIVES
INVOLVES

Intersectoral

Trunk System
Dam, main canal

Ma

Mb

Hc

MC

water

allocation
Waterlogging;
Salinization

DistributionSystem
Secondary& tertiary
canals

MM

Mb

HO

Me

Erosion

Introduction
of new
diseases

Terminal System (Onfarm)

Gravity
Sprinkler
Definitions:(H = high, M = moderate,L
See other secor tables.

EXTERNALITIES

H
H

M
H

L
L

L
L

L
L

low) USO = universalservice obligations

I The degree of subtractability

dependson scarcityof water resourcesand on maximumtransitcapacityof canals. It is assumedhere that subtractabilityincreasesat fartherpoints
from the initialwater intake.
' To some degree, this is a functionof both size and engineeringconfigurationof a given system. It is relativelyeasy to exclude farmsadjacentto large, lined canals from a
system,but far more problematicto excludeadjacentfarmers from a relativelysmall, unlinedcanal.
' Dependingon size and engineeringconfiguration.

oN

ANNEXII Page 5 of 16
SECTOR/SUBSECTOR:PORTS AND WATERWAYS
ACTIVITY

Piers, harbor
Ships
Port equipment
(oading/unloading)

NATURE OF GOOD/SERVICE
Degree of
Subtractability

Degree of
Excludability

M
H
H

H
H
H

PRODUCTION ASPECTS
Degree
of
Sunkness

Economies
of
Scale

H
L
M

M
L
L

EXTERNALITIES
OR

Degreeof
Coordination
Necessary

H
L
L

~~~~~~~~~~~~SOCIA&L
OBJECTIVES
INVOLVED

National
defense
concerns
Water Pollution
-J

Definitions: (H = high, M = moderate, L = low) USO = universal service obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all users. Low subtractability (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability
consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability = = it is relatively easy (i.e. non-costly) to prevent users from consuming it.
Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concerns,
etc.

ANNEXII Page 6 of 16
SECTOR(SUBSECTOR:RAILWAYS
ACTIVITY

Railbed
Switchingand Signaling'
Railcars
Freight Loading/Unloading
Definitions: (H = high, M = moderate, L

NATURE OF GOOD/SERVICE

PRODUCTION ASPECTS

Degree of
Subtractability

Degreeof
Excludability

Degree
of
Sunkness

M
L
H
H

H
M
H
H

H
L
L
M

T
Economies
of
Scale
M
L
L
L

Degree of
Coordination
Necessary

EXTERNALITES
OR
~~~~~~SOCL4L
OBJECTIVES
INVOLVED

H
H
H
M

USO
Network
externalities

low) USO = universal service obligations

Subtractability (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all users. Low subtractability (nonrivalry) =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability = = consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability = = it is relatively easy (i.e. non-costly) to prevent users from consuming it.
Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are rclatively eostless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control overthe rights to performrthe function, e.g., becauseoftechnicalconditions
etc.

forefficiency, need for equipment standards, safety concerns,

The more technically spocialized rail operations (e.g. high speed passenger, refrigerated freight) are less contestable and require greater coordination.

00

ANNEX II Page 7 of 16

SECTOR/SUBSECTOR:
ROADFREIGHTTRANSPORT
ACTIVITY

NATUREOF GOOD/SERVICE
Degreeof
Subtractability

PRODUCTIONASPECTS

Degreeof
Excludability

Degree
of
Sunk-

Economies
of
Scale

ness

Trucking Services

Degreeof
Coordination

EXTERNALITIES
OR
SOCIAL
INVOLVED

Necessary

Air, Noise
Pollution

Definitions: (H = high, M = moderate, L = low) USO = universal service obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all users. Low subtractbility (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability = = consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability

it is relatively easy (i.e. non-costly) to prevent users from consuming it.

Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concerns,
etc.

ANNEXII Page 8 of 16
SECTOR/SUBSECTOR: ROADS AND HIGHWAYS
ACTIVITY

NATURE OF GOOD/SERVICE
Degree of
Subtractability

Primary Roads
(National, Trunk)
SecondaryRoads
Tertiary roads - rural
urban
Signalingand traffic
control
Definitions: (H = high, M = moderate, L = low)

L
M
L
H
L

Degreeof
Excludability

Ma

Ma
Ma

L
L

PRODUCTION ASPECTS
Degree
of
Sunkness

H
H
H
H
L

Economies
of
Scale

L
L
L
L
L

Degree of
Coordination
Necessary

L
L
L
H
H

EXTERNALITIES
OR
SOCIAL
OBJECTIVES
INVOLVED

Affects
settlement
patterns,
drainage,
erosion,
public
safety, dust
and noise
pollution

USO = univeral service obligations

Subtractability (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of atl users. Low subtractability (nonrivalry) =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability = = consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability = = it is relatively easy (i.e. non-costly) to prevent users from consuming it.
Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perforn the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concerns,
etc.

a High for tottroads.

ANNEX II Page 9 of 16
SECTOR/SUBSECTOR: SOLID WASTE MANAGEMENT
ACTIVITY

NATURE OF GOOD/SERVICE

PRODUCTION ASPECTS

Degreeof
Subtractability

Degreeof
Excludabbiity

Degree
of
Sunkness

Economies
of
Scale

Degreeof
Coordination
Necessary

Collection

Transfer

Disposal
Sanitary land fill
Incineration

L
M

M
H

M
M

H
H

H
H

Resource Recovery
(recycling)

EXTERNALITIES
OR
OBJECTIVES
O LVES

Publichealth,
land, water,
pollution

Definitions:(H = high, M = moderate,L = low) USO = universalservice obligations

Subtractabilitv(or rivalry) refers to the impactthat consumptionby incrementalusers has on consumptionopportunitiesof all users. Low subtractability(nonrivalry) = =
consumptionby one user does not impedeavailabilityto other users (suchgoodsare jointlyconsumed).
High subtractability=
consumptionby one user imposesidentifiable,
ealculablecosts on other users (such goods are privately consumed).
Excludabilityrefers to the feasibilityof controllingaccess to a good. High excludability=

it is relativelyeasy (i.e. non-costly)to preventusers from consumingit.

Sunknessrefers to extent of sunk capital costs incurred in the event of ex.t from the activity. In activitieswith high contestability,entry and exit are relativelycostless(key
crkerion for contestabilityis the absence of sunk costs incurredin the event of exit).
Coordinationrefersto needfor controlover the rightsto performthe function,e.g., becauseof technicalconditionsfor efficiency,needfor equipmentstandards,safetyconerns,
etc.
a Some coordination
needed becauseof "economicsof contiguity' (withina given area, it is cheaper for adjacentpropertiesto be servedby one collectionservice agencythan
by multipleagence).

ANNEXII Page 10 of 16
SECTOR/SUBSECTOR:TELECOMMUNICATIONS- (BASIC NETWORK)
ACTIVITY

NATUREOF GOOD/SERVICE
_

Degreeof
Subtractability

__

__

__ __

__

__ __

Degreeof
Excludability

PRODUCTION ASPECTS
_

Degree
of
Sunk-

Economies
of
Scale

ness

Degreeof
Coordination

SOCIAL
OBJECTIVES

Necessary

NETWORK:
-Transmission
Local
Long Distance
-Switching

M
M
M

H
H
H

H*
L
L*

H*
L
L*

H
H
H

TERMINALEQUIPMENT:
-Individual
-common

H
M

H
H

L
M

L
L

M
M

Definitions: (H = high, M = moderate, L = low)

EXTERNALITIES
~~O
R

Network
effects,
USO

SO = service obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all use.
Low subtractability (nonrivalry) = =
consumption by one user does not impede availabilityto other users (such goods are jointly consumed). High subtractability = = consumption by one user imposes identifable,
edaulabk costs on other users (such goods armpnvatey consumed).
Excludability refers to the fessibility of controling access to a good. High excludability = = it is relatively easy (i.e. non-costly) to prevent users from consuming it.
Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concerns,
etc.

* Varies with technology

ANNEXII Page 11 of 16
t
SECTOR/SUBSECTOR: TELECOMMUNICATIONS- (NEIWORK EXTENSIONS)

ACTIVITY

NATURE OF GOOD/SERVICE
Degree of
Subtracta-

Degreeof
Excluda-

bility

bility

Value-AddedServices

Cellular, Paging,
MicrowaveRelay

___________________________

PRODUCTION ASPECTS

OR

Degree
of Sunk-

Degreeof
Coordination

ness

Scale

Necessary

Lb

~~~~~~~~~~~SOCIL4,
OBJECTIVES
INVOLVED

Networks

I~~~

I
=

Economies
of

Private or Specialized

Definitions:(H - high, M

EXTERNALIES

moderate,L

= low)

I__

.I~

~~~~~~

USo = univeral service obligations

Subtractability(or rivalry) refer to the impact that consumptionby incrementalusers has on consumptionopportunitiesof all users. Low subtractability(nonrivalry) =
consumptionby one user does not impedeavailabilityto other users (such goodsare jointly consumed). High subtractabihity consumptionby one user imposesidentifiable,
calculablecosts on other usen (such goodsarmprivately consumed).
Excludabilityrefersto the fcasibilityof controllingaccess to a good. High excludability=

it is relativelyeasy (i.e. nc:-costly) to preventusers from consumingit.

Sunknessrefcrs to extent of sunk capital costs incurredin the event of exit from the activity. In activitieswith high contestability,entry and exit are relativelycostless (key
criterionfor contestabilityis the absenceof sunk costs incurredin the event of exit).
Coordinationrefersto necd for controlover the rightsto perform the function,e.g., becauseof technicalconditionsfor efficiency,needfor equipmentstandards,safetyconcerns,
etc.

These usuallyare linkedto the basic networkfacilities.

Contctability dependson gettingaccss to mdio spectrum.

ANNEXII Page 12 of 16
SECTOR/SUBSECTOR:URBANTRANSPORT
ACFIVITY

PRODUCTION ASPECTS

NATURE OF GOOD/SERVICE
Degree of
Subtracta-

Degree of
Excluda-

Degree
of

Economies
of

Degreeof
Coordina-

bility

bility

Sunkness

Scale

tionI
Necessary

H
H

H
H

L
L

L
L

L
L

M
H
L

H
H
M

H
L
L

M
L
L

H
H
H

Van, bus (urban,

inter-urban)
Taxi
Metro/rapidtransit
- Tracks/Rails
- Rail cars
Traffic signaling
Definitions: (H = high, M = moderate, L = low)

EXTERNALITIES
OR
OBECTIAE
INVOLVED
OL

Air
Pollution,
USO
PublicSafety

USO = universal swrvice obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental user. has on consumption opportunities of all usrs. Low subtractability (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability = = consumption by one user imposes identifiable,
caculable costs on otier user. (such goods are privatdy consumed).
Excludability refers to the feasibility of controlling acness to a good. High excludability =

it is relatively easy (i.e. non-costly) to prevent users from consuming it.

Sunknes refers to extant of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit arcrelatively costless (key
crierion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refen to need for control over the rights to perforn the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concems,
dc.

ANNEX II Page 13 of 16
SECTOR/SUBSECTOR:
ACTIVITY

NATURE OF GOOD/SERVICE

PRODUCTION ASPECTS

EXTERNALITIES

OR
~~~~~~SOCALL

Degree of
Subtractability

Degree of
Excludability

Degree
of
Sunkness

Economies
of
Scale

Degree of
Coordination
Necessary

OBJECTIVES
INVOLVED

L
L
L

H
H
H

H
H
H

M
M
M

M
L
L

Public health
Water
pollution

Conventional
Street Sewer
Pumping Station
Treatment Plant

Definitions: (H = high, M = modcrate, L

WASTEWATER MANAGEMENT (CONVENTIONAL SEWERAGE)

low)

USO = universal service obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all users. Low subtractability (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability
consumption by one uscr imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability = = it is relatively easy (i.e. non-costly) to prevent users from consuming it.
Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the cvent of exit).
Coordination rcefrs to need for control over the rights to perform the function, e.g., because of technical conditions for efficicncy, need for equipment standards, safety concerns,
etc.

ANNEXI

Page 14 of 16

SECTOR/SUBSECTOR:WASTEWATER MANAGEMENT(INTERMEDIATE/LOW COST SEWERAGE)

ACTIVITY

NATURE OF GOOD/SERVICE
Degree of
Subtractability

Degreeof
Excludability

PRODUCTION ASPECTS
Degree
of
Sunk-

Economies
of
Scale

EXTERNALITIES
OR
SOCLAL
Degreeof
OBJECTIVES
Coordination
INVOLVED
Necessary

ness

CondominialSewerage
Localizedtreatment
Other intermediate
sewerage"
Basic sanitation(pit latrine)

Definitions: (H = high, M = moderate, L

M
M

M
M

M
M

L
L

Hb
HI

Ho

Publichealth
Water
pollution

low) USO = universal service obligations

Subtractability (or rivalry) refers to the imnpactthat consumption by incremental users has on consumption opportunities of all users. Low subtractability (nonrivalry)
=
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability
consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludabilitv refers to the feasibility of controlling access to a good. High excludability = = it is relatively easy (i.e. non-costly) to prevent users from consuming
it.
Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless
(key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety concerns,
etc.

I Includes approaches such as solids-free sewerage, simplified, shallow, and flat grade sewerage.
b Condominial sewerage requires coordination among neighboring property owners.

ANNEXII Page 15 of 16
SECTOR/SUBSECTOR:

ACTIVITY

WATER SUPPLY (NONPIPED)

NATURE OF GOOD/SERVICE
Degreeof
Degree of
SubtractaExcludability
bility
.____________I

PRODUCTION ASPECTS
Degree
of
Sunkness

Economies
of
Scale

Degreeof
Coordination
Necessary

O CIVES
INVOLVED

Public
health
Depletionof
Aquifer

VendorTanks

Borehole

Defrnitions: (H = high, M = moderate, L = low)

EXTERNALITIES
OR

USO = universal service obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental users has on consumption
opportunities of all users. Low subtractability (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed).
High subtractability = = consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludability refers to the feasibility of controlling access to a good. High excludability =

it is relatively easy (i.e. non-costly) to prevent users from consuming it.

Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities
with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., becauseof technical
conditions for efficiency, need for equipment standards, safety concerns,

ANNEXII Page 16 of 16
SECTOR/SUBSECTOR:WATER SUPPLY (PIPED)
ACTIVITY

PRODUCTION ASPECTS

NATURE OF GOOD/SERVICE
Degree of
Subtractab clity

Degreeof
Excluda-

Degree
of

Economies
of

Degree of
Coordina-

bility

Sunkness

Scale

tion
Necessary

Trunk System (intake

pumpingstation)

DistributionSystem

Terminal Equipment:
Common(i.e. handpump)
Individual(i.e. home

faucet)

Definitions: (H = high, M = moderate, L

H
H

H
H

H
H

EXTERNALITIES
OR
SOC[AL
OBVOLVED
ED

nWOL'

Intersectoral
resource
allocation
Public
health
o
USO

low) USO = universal service obligations

Subtractabilitv (or rivalry) refers to the impact that consumption by incremental users has on consumption opportunities of all users. Low subtractability (nonrivalry) = =
consumption by one user does not impede availability to other users (such goods are jointly consumed). High subtractability
consumption by one user imposes identifiable,
calculable costs on other users (such goods are privately consumed).
Excludabilitv refers to the feasibility of controlling access to a good. High excludability =

it is relatively easy (i.e. non-costly) to prevent users from consuming it.

Sunkness refers to extent of sunk capital costs incurred in the event of exit from the activity. In activities with high contestability, entry and exit are relatively costless (key
criterion for contestability is the absence of sunk costs incurred in the event of exit).
Coordination refers to need for control over the rights to perform the function, e.g., because of technical conditions for efficiency, need for equipment standards, safety coneems,
etc.

ANNEX 111 - PREVALENCE OF CERTAIN FORMS OF

PRIVATE SECTOR PARTICIPATION IN INFRASTRUCTURE BY SECTOR

ANNEXIII Page I of 7

PRIVATE SECTOR PARTICIPATION IN IN1RASTRUCTUREBY SECTOR

lnfrastrudure TyppjNOT/
Region/Country

1100

coDes
moo

Servke
Contrcts

.
Leasg

MIhagement
Contracts

Bonds

POWER
Afrks
Cote d'lvoire

lbe Gambia
Guinea-Bisau

P
0
P

Siem Leonc

AAh
0, P

China

HongKong
India

nldonesia
Malaysia
Pakistan

I
I
I

O. P

0
0

0.1

PhfyWines

Tbailad
Vietnam

EA^,

hbEd-Rat,N. Afrka
P

ey

Joidan
Onu
Polam

Tuaey

I, P

Lash Amneric& Caibbamu

Argaina-

Barb,ados

Bdolivia

0
Key:

P - Pbuked

I -Initisd

Opendonal

ANNEX III Page 2 of 7

Wrastructure Type'
Region/Country

BOT/
BOO

Concession

Leasing

Service
Contracts

Management
Contracts

Bonds

POWER
(continued)
Brazil

Colombia

0
0

DominicanRepublic

Ecuador
Honduras

Venezuela

TRANSPORT

Benin
BurkinaFaso

Bufundi

Cameroon

Car

Cape Verde
P

Chad

CongoB.

Ethiopia
I

Gambia

Ghana

Key

P =Planned

I =Initiated

0= Operational

ANNEX III Page 3 of 7

BOTIBOO

Infrastructure Type/
Region/Country

Concession

Leasing

Service
Contracts

Management
Contracts

TRANSPORT
(Continued)
Guinea
Guinea-Bissau

Kenya

Lesotho

Madagascar

Malawi

Mali

Mauritius

Mozambique

Niger

Nigeria
0

Senegal
I

Tanzania
Togo

I
I

UgandaI

Zaire

Asia

Bangladesh
I

China

Hong Kong

O,I,P

India

O,P

Indonesia
Key:

P = Planned

I = Initiated

0 = Operational

Bonds

ANNEX HI Page 4 of 7

Iafriniructuv Type/
RaC
ntzy

DOT/
BOO

C_me.-

Sw,ke
Ceabds

urea

Manag-__t
Cenfts

B_on

TRANSPORT
rama

Malyia

O,P

Nepal

Pakidan

Papua NewGuu a

hwippis

IP

Thailaud

LP

Empe, MiEast,

N. Afics

Algea

~~~~

Egypt
Jordan

~~0

moucco

o
P

Tunisa

O,P

Ladin Ams*a & CArilb.m

ArgeiOi-

Brazil

Chile

Colom,bia

1,0

Mexico

0
Key

P - Planned

Initiewtd

0 - Opeioal

ANNEX HI Page 5 of 7

| Irastncture, Type/
Region/Country

BOTI
BOO

Concession

Leasing

Servie
Contracts

Management
Contracts

Bonds

WATER & SEWERAGE

MAh

Cote d'lvoire

Guinea

Guinea-Bi.aau

Lesotho

Rwanda

Sao ToMe

'Me Ganbia

Asia

Bangladesh

China

Hon Kong

India

Indonesia

Macau

Malaysia

Pakisan

Philippines

Thailand

P
Key

Planned

I = Initiated

-Operational

ANNEX m Pap 6 of 7

h
R

li
Type/
C300 yBOOSm

DOTI

CAc.

Solic
CLssiacts

_
C J_acs

Dm

WATER& SEWERAGE
Eup,

MNiNEnAd,
N. Africa

Egy

P,O

Joedkn

Mrocco

Yanen

JAh

Ameica & Carbb.m

Argmi

floivi

Brazil

Cbile

Colombia

ldexco

Veneal

0
0

I
I

TELECOMMVMCATIONS
Ads
bIdia

Philippine.

hailand

P
Ky

P - Planned

I - Inad

0-

Opeational

ANNEXmI Page 7 of 7

b_uctwe
RCmu~y

Typ/

DOT/

BO

Senke
im

LeCont

Maonatm t
CmCcts

B_m

TELECOMMUNICAnONS

Dolivia

ken!

COlOadiaI

IIUUGATION

Nigad

Ask
Indomai

EKee, MiW-Endt,N. AHa

Uorocco

KCey

P -Phlanncd

,, I

I -Init.iated

0 - Operatic..1

76
REFERENCES

Augenblick, Mark and B. Scott Custer, Jr. 1990. 7he Build, Operate, and Transfer ("BOT") Approach
to Infrastructure Projects in Developing Countries. Policy, Research and External Affairs
Working Paper No. 498. World Bank, Washington, D.C.
Bartone, Carl R. 1991. "Institutional and Management Approaches to Solid Waste Disposal in Large
Metropolitan Areas", Waste Managing & Research, 9, 525-536.
Bartone, Carl R. et al. 1991. "Private Sector Participation in Municipal Solid Waste Service: Experiences
in Latin America." Waste Management and Research, 9, 495-509.
Baumol, William J., Panzar, John C., and Robert D. Willig. 1988. Contestable Markets and the Theory
of Industry Structure. San Diego: Harcourt Brace Jovanovich.
Bennathan, E., J. Gutman, and L. Thompson. 1991. Reforming and Privatizating Poland's Road Freight
Industry. Infrastructure and Urban Development Department Working Paper No. 750. World
Bank, Washington, D.C.
1992. Reforming and Privatizing Hungary's Road Haulage. Infrastructure and Urban Development
Department Working Paper No. 790. World Bank, Washington, D.C.
Berthier, Jean. 1991. French Motorways Operated by ConcessioniaryCompanies: Historical Back-ground
and Current Situation. Bureau Central Pour l'Equipement d'Outre-Mer. Paris, France.
Besant-Jones, John E., ed. December 1990. Private Sector Participation in Power thru BOOT Schemes.
Energy Series Paper No. 33. World Bank, Industry and Energy Department, Washington. D.C.
Campbell, Tim., et al. 1991. Decentralization to Local Government in LAC. National Strategies and
Local Response in Planning, Spending and Management. Regional Studies Program Report No.
5. World Bank, Latin America and the Caribbean Technical Department, Washington, D.C.
Carbajo, Jose, ed. 1993. Regulatory Reform in Transport. Some Recent Experiences. A World Bank
Symposium. World Bank, Washington, D.C.
Cellier, Jacques. 1991 Decentralizing of Road Administration. Infrastructure and Urban Development
Department Infrastructure Notes No. RDI 1. World Bank, Washington, D.C.
Cointreau-Levine, Sandra. 1992. Private Sector Participation in Mun2icipalSolid Waste Services in
Developing Countries. Forthcoming in Urban Managment Program Working Paper Series. United
Nations Developmnent Program (UNDP)/World Bank/United Nations Centre for Human
Settlements, Washington, D.C.

77

Coyaud, Daniel. 1988. Private and Public Alternativesfor Providing Water Supply and Sewerage
Services. Infrastructire and Urban DevelopmentDepartmentReport No. INU-31. World Bank,
Washington.D.C.
Dagerman, Lo. 1992. "Deregulationof Local Busesin the U.K. and Santiago-Some Lessons." World
Bank, Infrastructureand UrbanDevelopmentDepartment,TransportationDivision, Washington,
D.C.
Davey, Kenneth. 1988.MunicipalDevelopmentFundsandIntermediaries.Policy, Planningand Research
WorkingPaper 32. World Bank, Washington,D.C.
Donohue, John D. 1989. The PrivatizationDecision:Public Ends, Private Means. New York: Basic
Books, Inc.
Electricite de France International. May 19, 1992. "Power Utility Management by Performance
Contracting." Presentationat WorldBank, Washington,D.C.
Fox, William F. 1992. "Strategic Options for Urban Infrastructure Management." World Bank,
Infrastructureand Urban DevelopmentDepartment,Washington,D.C.
Galal, Ahmed, LeroyJones, ParkajTandor, and Ingo Vogelsang.WelfareConsequensesof SellingPublic
Enterprises.World Bank, Washington,D.C. (Forthcoming)
Galenson, Alice and Louis Thompson. 1992. The Bank's Evolving Policy Toward Railway Lending.

World Bank DiscussionPaper (Forthcoming).Transportation,Water and Urban Development

Department,Washington,D.C.
Galenson, Alice. 1993. "Private Sector Development in Transportation: Railways." World Bank,
Transportation,Water and Urban DevelopmentDepartment,Washington,D.C. Draft, April.
1990. New Zealand: Road OperatedLike a Public Utility.Infrastructureand Urban Development
DepartmentInfrastructtireNote No. RD-I. World Bank, Washington,D.C.
1989. Labor Redundancyin the TransportSector. Policy, Planningand ResearchWorkingPaper
No. 158. World Bank, Washington,D.C.
Gyamfi, Peter and others. 1992. InfrastructureMaintenancein LAC 7he Costs of Neglectand Options
for Improvement,RegionalStudiesProgram. ReportNo. 17. WorldBank, Latin Americaand the
CaribbeanTechnical Department,Washington,D.C.
Heggie, Ian G., and Quick, Michael. 1990. ImprovingManagementand ChangingPoliciesfor Roads:
An Agendafor Reform. Infrastructureand Urban DevelopmentDepartmentReport No. INU 92.
World Bank, Washington,D.C.
Hilton, Rita. 1992. "Participationand Delivery of InfrastructureServices." WorldBank, Infrastructure
and Urban DevelopmentDepartment,Washington,D.C. Draft, September.

78
1992. "Legal and Regulatory Framework: Integral Components of Institutional Arrangements for
Delivery of Infrastructuire Services." World Bank, Infrastructure and Urban Development
Department. Washington, D.C. Draft, September.
1991. "Institutional Arrangements for the Provision and Production of Infrastructure." World Bank,
Infrastructure and Urban Development Department, Washington, D.C. Draft.
Humplick, Frannie and Samer Madanat. 1992. "Household Respones to Reliability of Water Supply: The
Case of Istanhul. Turkey and Faisalabad, Pakistan." World Bank, Infrastructure and Urban
Development Department, Water and Sanitation Division, Washington, D.C.
Israel, Arturo. 1992. Issuesfor Infrastructure Management in the 1990s. World Bank Discussion Paper
No. 171. Washington. D.C.
Kikeri, Sunita, John Nellis and Mary Shirley. 1992. Privatization: 7he Lessons of Experience.
Washington, D.C.: World Bank.
Lethbridge, John and Zvi Ra'anan. 1991. Port Administration: Public vs Private Sector. Infrastructure
and Urban Development Department Infrastructure Notes, No. PS-5. World Bank, Infrastructure
and Urban Development Department, Washington, D.C.

Levy, Brian and Pablo Spiller. 1993.Regulation,Institutions,and Committmentin Telecommunications:

A Comparative Analysis of Five Country Studies. Paper presented to World Bank Conference on
Development Economics, May 3-4, 1993. World Bank, Washington, D.C.
Levy, Hernan and Aurelio Menendez. 1992. Privatization in Transport: The Case of Port Kelang
(Malaysia) ConitaiinerTerminal. Economic Development Institute Working Paper. World Bank,
Washington, D.C.

1992. Privatizationin Transport:Contestabilityin Railway Operationsfor the FertilizerIndustry

in Senegal. Economic Development Institute Working Paper Catalogue No. 705/027. World
Bank, Washington, D.C.

1990. Privatizationin Transport.ContractingOut the Provisionof PassengerRailwayServicesin

Thailand.Econoinic Development Institute Working Paper Catalogue No. 705/023. World Bank,
Washington. D.C.
Lorrain, Dominique. 1992. "The French Model of Urban Services." West European Politics, Vol. 15,
No.2, pp. 77-92.
Miquel, Sergio and James Condron. 1991. Assessment of Road Maintenance by Contract. Infrastructure
and Urban Development Department Report No. INU 91. World Bank, Washington, D.C.
Moser, Carolyn .O.N.. "CommiiunityParticipation in Urban Projects in the Third World," Progress in
Planning 32 (1989): 73-134.

79
Moyer, Neil E. and Louis S. Thompson. 1991. Optionsfor Reshapingthe Railway. Policy Research
Working Paper No. 926. World Bank, Infrastructure and Urban DevelopmentDepartment,
Washington.D.C.
Narayan, Deepa. 1993. "The Challengeof ParticipatoryDevelopment:LessonsLearned in Rural Water
Supply." World Bank. Water and SanitationDepartment,Washington,D.C. Draft, March.
Nellis, John. 1988. ContractPlans and Public EnterprisePerformance.Policy, Planningand Research
WorkingPaper No. 118. World Bank, Country EconomicsDepartment,Washington,D.C.
Paul, Samuel. 1991. Accountabilityin Public Services:.Exit, Voice and Capture. Policy, Research and
External AffairsWorkingPaper No. 614. World Bank, Country EconomicsDepartment,Public
Sector Managementand Private SectorDevelopmentDivision, Washington,D.C.
Scurfield, Richard G. 1992. "Developing the Role of the Private Sector in Ports." World Bank,
Infrastructureand UrbanDevelopmentDepartment,TransportDivision,Washington,D.C. Draft.
1990. CompetitiveTenderingFor Public Bus Services. Infrastructure and Urban Development
DepartmentInfrastructureNote No UT-I. World Bank, Infrastructureand Urban Development
Department,Washington,D.C.
Sherer, F.M. and David Ross. 1990.IndustrialMarket Structureand EconomicPerjbrmance.Boston:
Houghton Mitflin.
Silverman, Jerry M. 1990. Public Sector Decentralization.Economic Policy Reform and Sector
Investment Programs. Public Sector ManagementDivision Study Paper No. 1. World Bank,
Africa Technica!Department,Washington,D.C.
Tanahashi, Yasushi. 1991. "Reform of Railways in Japan." Infrastructure and Urban Development
Department Discussion Paper 99, World Bank, Infrastructure and Urban Development
Department,Washington,D.C.
Triche, Thelma A. 1992. "PSP in Urban Water Supply: Issues, Implicationsand Examples." World
Bank, Infrastrtictureand Urban DevelopmentDepartment,Washington,D.C. Draft.
1991. "Seminar on Privatizationof Water Supply in the U.K. and The Role of the Private Sector
in France. Summary of The Two Approaches and Discussion of Issues."World Bank,
Transportation. Water and Urban DevelopmentDepartment, Water and SanitationDivision,
Washington,D.C.
1990. Private Participationin the Delivery of Guinea's Water Supply Services. Policy, Research
and External Affairs Working Paper No. 477. World Bank, Infrastructure and Urban
DevelopmentDepartment,Washington,D.C.
1990. Private Participationlin Water Supply. Infrastructureand Urban DevelopmentDepartment
Infrastructure Note No. WS-1. World Bank, Washington,D.C.

80
United States Agency tor InternationalDevelopment(USAID), Engineering-Science,Inc. (Pasadena,
California)and ResouircesDevelopmentConsultants,Ltd. (Sri Lanka). 1991.ExecutiveSummary
Final Report on Instititional Developmentof the National Water Supply and Drainage Board.
Washington.D.C.
United States Environmental Protection Agency (EPA). 1990. Public Private Partnershipsfor
Environmental Facilities: A Self-Help Guide for Local Governments. Washington, D.C.:
GovernmentPrintingOffice.
Wellenius, Bjorn and others. 1993. Telecommunications:WorldBank Experienceand Strategy. World
Bank DiscussionPaper No. 192. World Bank, Washington,D.C.
Whittington, Dale et al. 1992. 'Possible Adverse Effects of Increasing Block Tariffs in Developing
Countries." EcoinomicDevelopmentand CulturalChange 41, no. 1: 75-87.
World Bank. 1991. 7he Reform of Public Sector Management:Lessonsfrom Experience. Policy and
Research No. 18, World Bank, Policy, Research and External Affairs, Country Economics
Department,Washington,D.C.
Yates, Philip. 1988. Improving Power System Efficiency in the Developing Countries 7hrough
Performance Contracting. Energy Series Working Paper No. 4. World Bank, Industry and

EnergyDepartment,Washington,D.C.
Yepes, Guillermo. 1990. Mangementand OperationalPracticesof Municipaland Regional Waterand
SewerageCompaniesin Latin Americaand the Caribbean.Infrastructureand UrbanDevelopment
DepartmentReport No. INU 61. World Bank, Washington,D.C.

81

OTHERSOURCES
The following are internal World Bank documents that may not be available to the public:

World Bank. 1993. "Municipal Development Sector Review." World Bank, Transportation, Water and
Urban Development Department, Urban Development Division, Washington, D.C. Draft, June.

Distributors of World Bank Publications

ARGENTINA
Carlos Hirsch, SRL
Galerie Guemes
Florida 165,4th Floor-Ofc. 453/465
1333Buenc Air
AUSTRALIA, PAPUA NEW GUINEA,
FiJI,SOLOMONISLANDS,
VANUATU, AND WESTERN SAMOA
DA. Books & journals
648Whitehorse Road
Mitcham 3132
Victoria

TheMiddleEast Observer
41, Sherif Street
Cairo
FINLAND
Akatemeinen lrjakauppa
P.O Box 128
SF-{0101 Helsinki 10
FRANCE
World Bank Publications
66,avenued'Ih
75116Paris

AUSTRIA
Gerold and Co.
Graben 31
A-1011 Wien

GERMANY
UNO-Verlag
Poppelsdorfer Allee 55
D-5300 Bonn I

BANGLADESH
Micro Industries Development
Assstance Society (MIDAS)
HouseS, Road 16
Dhanmondi R/Area
Dhaka 1209

HONG KONG, MACAO

Asi2000 Ltd.
46-48Wyndham Street
Winning Centre
2nd Floor
Central Hong Kong

Bnch eficn:
156, Nur Ahmed Sark
Chittagong 4000
76 K.D.A. Avenue
Kulna 9100
BELGIUM
JeanDe Lannoy
Av. du Ri 202
1060Brussels
CANADA
Le Diffuseur
CP. B5,1501B rue Ampere
Boucherville, Qutbec
J4BSE6
CHILE
Invertec IGT SAAmerico VespurioNorte 1265
Santiago
CHINA
China Financial & Economnu
Publishing House
8, Da FoSi Dong Jie
Beijing
COLOMBIA
InfoenlaceLtda.
ApartadoAereo34270
BogotaD.E.
COTE DIVOIRE
Centre d'Edition et de Diffusion
Africaines(CEDA)
04 B.P. 541
Abidjan 04 Plateau
CYPRUS
Centerof Applied Research
Cyprus College
6, DiogenesStreet, Engomi
P.O.Box 2006
Nicosia

INDIA
Allied Publishers Private Ltd.
751Mount Road
Madras- 600002
Bnnch oe7les:
15JN. Heredia Marg
Ballard Estate
Bombay - 400038
13/14 Asaf Ali Road
New Delhi -110002
17Chittaran;an Avenue
Calcutta - 700 072
JayadevaHostel Building
5th Main Road,Gandhinagar
Bangalore-560009
3-5-1129Kachiguda
Cron Road
Hyderabad -500027
Prarthan Flat, 2nd Floor
Near Thakore Baug, Navfangpura
Ahmedabad - 380009
Patial House
16-A Ashok Marg
Lucknow-226001
Central BazaarRoad
60 BajajNagar
Nagpur 440010
INDONESIA
Pt Indira Limited
JalanBorobudur 20
P.O.Box 11
Jakarta 10320
IRELAND
Govemment Supplies Agency
4-5 Harcourt Road
Dublin 2

DENMARK
SamfundsUtteratur
Rosenoemns
All 1
DK-1970 FrederiksbergC

ISRAEL
Yormot Uterature Ltd.
P.O.Box 56055
Tel Aviv 61560

DOMINICAN REPUBLIC
Editora Taller, C por A.
Restauraci6neIsabel la Catica 309
ApartadodeCorreos2190Z-1
Santo Domingo

ITALY
Licosa Commimionaria SansoniSPA
Via Duca Di Calabria,1/1
CaelaaPostale552
50125Firenze

EGYPT,ARABREPUBUCOF
Al Ahram
Al Galaa Street
Cairo

JAPAN
Easten Book Service
Hongo S-Chome, Bunkyo-ku 113
Tokyo

KENYA
Africa Book Service(EA) Ltd.
Quaran House, MfanganoStrar
P.O.Box 45245
Nairobl
KOREA, REPUBLIC OF
Pan Korea Book Corporation
P.O Box 101,Kwangvwhamun
Seoul
MALAYSIA
University of Malaya Cooperative
Bookahp, Limiited
P.O.Box 1127,JaianPantai Baru
59700 Kuala Lumpur
MEXICO
INFOTEC
Apartado Postal 22460
14060Tbalpan.Mexico D.F.
NETHERLANDS
De Lindeboom/InOr-Publikaties
P.O.Box 2C2
7480 AE Haaksberpen
NEWZEALAND
EBSCONZ Ltd.
Privte Mail Bag 99914
New Market
Auckland
NIGERIA
University Pres Limited
Three Crowns Building Jerxcho
PrivateMad Bag 509S
Ibadan
NORWAY
Nerves Information Center
Book Department
P.O.Box 6125 EAterstad
N-0602Oslo 6
PAKISTAN
Mirn Book Agency
65, Shahrah-.Quaid-e-Azam
P.O.Box No 729
LAhore54000
PERU
Editorial Desarrollo SA
Apartado 3824
Lima 1
PHfUPPINES
Internationl Book Center
Suite 1703.CityLand 10
Condominium Tower I
Ayala Avenue, Corner H.V. dela
Costs Extension
Makati, Metro Manila
POLAND
International Publishing Service
Ult Pfekna31/37
00-677Warzawa
For nshcr*ttm ordes:
IPSJounals
Ul Okezna3
02-9l6 Warszawa
PORTUGAL
Livmria Portugzl
RuaDoCarmo70-74
1200Lisbon
SAUDI ARABIA, QATAR
Jarir Book Store
P.O.Box3196
Riyadh 11471
SINGAPORE.TAIWAN,
MYANMARLBRUNEI
Infortiabon Publications
Private, Ltd.
Golden Wheei Building
41. Kallang Pudding, o04-03
Singapor 1334

SOUTH AFlICA, BOTSWANA

Forsingle title:
Oxford Univerity Prm
Southern Africa
P.O. Box 1141
Cape Town 8000
Forsukcr*vte oers:
International Subscription Service
P.O. Box41095
Craighall
Johannesburg2024
SPAIN
Mundi-Prens Lbros, SHA
Caello 37
28001Madrid
Librera Internecional AEDOS
Consell de Cent 391
08009 Barcelona
SRI LANKA AND THE MALDIVES
Lake House Bookshop
PO Box 244
100,Sir Chittampalan A.
GardinerMawatha
Colombo2
SWEDEN
Forsingle tfitle
Fritzes Fackboksforetaget
Regeringatan 12,Box 16356
5-103 27 Stockholm
Foraarrcr*ionu oAri:
Wennergren-WlaliamsAB
P. 0. Box 305
S-17125 Solna
SWITZERLAND
For single t11w
Librairie Payot
1,rue de Bourg
CIH 1002 Lausanne
Foriubscr>tioorne:
Libraine Payot
Service des Abornnernents
Ca postale 3312
CH 1002Lauann
TANZANIA
Oxford University Preo
P.O. Box S299
Maktaba Road
Dare Salaam
THAILAND
Central Department Store
306Silom Road
Bangkok
TRINIDAD & TOBAGO, ANTIGUA
BARBUDA, BARBADOS,
DOMINICA, GRENADA. GUYANA,
JAMAfCA. MONTSERRAT, ST.
KITTS & NEVIS, ST. LUCIA.
ST. VINCENT & GRENADINES
Systemtics Studies Unit
#9 Watts Street
Cur"pe
Trinidad, WentInd ies
TURKEY
Infot l
Narlabahte Sok. No. 15
Cagaoglu
Istanbul
UNITED KINGDOM
MicroinfoLtd.
P.O. Box3
Alton, Hampshire CU34 2PG
England
VENEZUELA
Libreria del Este
Aptdo. 60337
Caracas106-A

Recent

World

Bank

Discussion

Papers

(continued)

No. 183

Participatory
Developmentand the WorldBank: PotentialDirectionsforChange.Bhuvan Bhatnagar and Aubrey C.
Williams, editors

No. 184

Agricultural
Researchin SouthernAfrica:A FrameworkforActiotn.Andrew Spurling, Teck Y. Pee, Godwin
Mkamanga, and Christopher Nkwanyana

No. 185

MilitaryExpenditureand EconomicDevelopment:A Symposiumon ResearchIssues.Edited by Geoffrey Lamb with

ValerianaKallab

No. 186

Efficiencyand Substitutionin Pollutiotn

Abatement:ThreeCaseStudies.Dennis Anderson and William Cavendish

No. 187

The StateHoldingCotnpatny:Issuesand Options.Anjali Kumar

No. 188

IndigenousViewsof Land and the Environment.Shelton H. Davis, editor

No. 189

Poverty,Population,and the Environment.Stephen D. Mink

No. 190

NaturalGas in DevelopingCountries:Evaluatinig
the Benefitsto the Environment.John Homer

No. 191

AppropriateMlacroecotiomic
Managementin Indonesia'sOpen Econotny.Sadiq Ahmed

No. 192

Telecomtnunications:
World Bank Experienceand Strategy.Bjom Wellenius and others

No. 193

InfornationSystemsStrategiesforPublicFinancialManagement.Hywel M. Davies, Ali Hashim, and Eduardo Talero

No. 194

SocialGainsfromFemaleEducation:A Cross-National
Study. K. Subbarao and Laura Raney

No. 195

Towardsa SustainableDevelopment:The Rio dejaneiroStudy. Edited by Alcira Kreimer, Thereza Lobo, Braz
Menezes, Mohan Munasinghe, and Ronald Parker

No. 196

EasternEuropein Transition:FromRecessionto Growth?:Proceedings

of a Conference
on the Macroeconomic
Aspectsof
Adjustment, Cosponsored
by the International
MonetaryFundand the World Bank. Edited by Mario 1. Blejer,
Guillermo A. Calvo, Fabrizio CoriceDli,and Alan H. Gelb

No. 197

KoreanIndustrialPolicy:Legaciesof the Pastand Directionsfor

the Future.Danny M. Leipziger and Peter A. Petri

No. 198

ExportingHigh-ValueFoodComnmnodities:
SuccessStoriesfrom DevelopingCountries.Steven M.Jaffee with the
assistanceof Peter Gordon

No. 199

Borrower
Ownership
ofAdjustmentPrograms
and thePolitical
EconomyofReform.John H.Johnson and SulaimanS. Wasty

No. 200

SocialIt!rastructureConstructiotn
in the Sahel:OptionsforImprovingCurrentPractices.
Bemard Abeille and
Jean-Mane Lantran

No. 201

Urbanization,Agricultural
Development,and Land Allocation.Dipasis Bhadra and Ant6nio Salazar P. Brandlo

No. 202

KoreanIndustrialPolicy:Legaciesof the Pastand Directionsfor

the Futjre. Danny M. Leipziger and Peter A. Petri

No. 203

PovertyReductionitnEastAsia: The Silent Revolutiorn.

FridaJohansen

No. 204

Managingthe Civil Service:The Lxssonsof Reformin IndustrialCountries.Barbara Nunberg

No. 205

Designinga Systemof LaborAlarketS.atisticsand Information.Robert S. Goldfarb and Arvil V. Adams

No. 206

InfornatiotnTechitiology
in WorldBank Lending:Increasingthe DevelopmentalImnpact.
Nagy Hanna and Sandor Boyson

No. 207

Proceedings
of a Conferenceon CurrencySu'stitution and CurrencyBoards.Edited by Nissan Liviatan

No. 208

DevelopingEffectiveEmploymentServices.David Fretwell and Susan Goldberg

No. 209

EvolvingLegalFrameworksforPrivateSectorDevelopmentin Centraland EasternEurope.Cheryl W. Gray and

Associates

No. 210

Marine Biotechnology
and DevelopingCountries.Raymond A. Zilinskas and Carl Gustaf Lundin

No. 211

RevitalizingAgricultural
Researchin the Sahel:A ProposedFrameworkforAction.Jan Weijenberg, Josue Dione,
Michael Fuchs-Carsh, Adolphe Ker, and Jacques Lefort

The World Bank

Headquarters
1818 H Street, N.W.
Washington, D.C.

European Office
66, avenue d'Frna
75116 Pavns,France

Telephone: (i02) 477-1234

Facsimile:(202) 477-6391
Telex: wiUg64145 WORLDBANK
RCA 248423 WORLDBK
Cable Address: INTBAFRAD

Telephone: (1) 40.69.30.00

Facsimile:(1) 40.69.30.66
Telex: 640651

Tokyo Office
Kokusai Building3co
1-1 MaBunouchi3-chome
Chiyoda-ku, Tokyo 100japan
Telephone: (3) 3214-5001
Facsimile: (3) 3214-3657
Telex: 26838

WASHINGTONDC

URD 100
12627
0-8213-2627-9
INSTITUTIONALOPTIONSIN

1111111111111
~111111111
400000009469

$7 .95

ISBN 0-8213-2627-9