UNFCCC/CCNUCC CDM – Executive Board Page 1

Version 01/ 02 August 2011

F-CDM-RtB

Name of the stakeholder1 submitting this form (individual/organisation):

Carbon Africa Limited

Matthew Woods (Director)

Address and Contact details of the individual submitting this Letter:

Address: P.O. Box 14938-00800, Nairobi, Kenya

Telephone number: +254706374150 or +254721620803

E-mail Address: info@carbonafrica.co.ke

Title/Subject (give a short title or specify the subject of your submission)

Recommending Changes to: Draft revision to the approved consolidated baseline and monitoring methodology ACM0016 “Mass Rapid Transit Projects”

Please mention whether the Submitter of the Form is:

Project participant

Other Stakeholder, please specify Project Developer

Specify whether you want the Letter to be treated as confidential2):

To be treated as confidential

To be publicly available (UNFCCC CDM web site)

Purpose of the Letter to the Board: Please use the space below to describe the purpose for submitting Letter to the Board.

(Please tick only one of the four types in each submission )

Type I:

Request Clarification Revision of Existing Rules Standards. Please specify reference

Procedures. Please specify reference

Guidance. Please specify reference

Forms. Please specify reference

Others. Please specify reference Draft revision to the approved consolidated baseline and monitoring methodology ACM0016 “Mass Rapid Transit Projects” (Version 03.0.0)

Type II: Request for Introduction of New Rules

Type III: Provision of Information and Suggestions on Policy Issues

1 Note that DNAs and DOEs shall not use this form to submit letter to the Board. 2 Note that the Board may decide to make this Letter and the Response publicly available

CDM: FORM FOR SUBMISSION OF “LETTER TO THE BOARD” (Version 01.1)

(To be used only by the Project Participants and other Stakeholders for submitting Letter to the Board as per Modalities and Procedures for Direct Communication with

Stakeholders)

UNFCCC/CCNUCC CDM – Executive Board Page 2

Version 01/ 02 August 2011

Please use the space below to describe in detail the issue that needs to be clarified/revised or on which the response is requested from the Board as highlighted above. In doing this please describe the exact reference source including the version (if any).

Carbon Africa Limited

P.O. Box 14938

00800, Nairobi

Kenya

Tel: +254 721 620 803 or

+ 254 706 374 150

Email: info@carbonafrica.co.ke

Website: www.carbonafrica.co.ke

24 November 2011

To: CDM Executive Board

UNFCCC Secretariat

Martin-Luther-King-Strasse 8

D 53153 Bonn

Germany

Cc: CDM Methodology Panel

Honorable Members of the CDM Executive Board, Dear Mr Hession, Revision of ACM0016 Baseline and Monitoring Methodo logy for Mass Rapid Transit Projects Carbon Africa Limited is a Kenya-based CDM consultant. We are advising transport sector company Bombardier Transportation on the realization of CDM projects in developing countries. Further to both Bombardier Transportation and our own previous inputs regarding the revision of ACM0016 and the Methodology Panel’s deliberations on the matter, Carbon Africa believes that the revised draft of the methodology could still use improvement in the following two areas: 1. The new approach to additionality is very much welcome. However, the proposed benchmark value (10%) regarding the contribution of CER revenue to operations and maintenance costs is not based on any transparent rationale and will restrict the application of ACM0016. We reiterate the evidence provided during the recent call for inputs and we refer again to a 2001 World Bank study, which shows that operating costs per passenger for rail systems range from USD 0.61 in Hong Kong to USD 0.19 in Santiago. We alert the EB that with these values CER revenues based on historic prices will rarely exceed 5% of operations and maintenance costs. Therefore we urge the EB to take into account the information provided by stakeholders in determining what the benchmark should be.

UNFCCC/CCNUCC CDM – Executive Board Page 3

Version 01/ 02 August 2011

2. Related to the first point, the approach to determine the expected CER revenue by using the average secondary CER price of the year prior the start of validation requires further consideration. In our opinion, such an approach does not account for (a) the forward looking perspective of investors and also (b) future inflation, which will raise project operational costs whereas historic CER prices do not incorporate this factor. To substantiate our argument, we would like to point out that the current secondary market price for CERs is around EUR 6.00, whereas the expected price for a post-2012 issued CER, (when new MRTS projects will become operational) is expected to have a lower range of EUR 15-17. Therefore we suggest to include a standardized option in the methodology for the determination of future CER price by allowing the project proponent to procure price forecasts for the year in which the project is first expected to generate CERs from three different independent analysts, the most conservative price of which will be used. Furthermore, there are areas of the draft revised methodology that may require editing: - On page 3, the applicability criteria still requires that the MRTS must replace existing bus routes, whereas the Methodology Panel 53rd meeting report indicates in paragraph 22 that this applicability criterion has been removed. - On page 49, there is still reference to year 7 of the crediting period. As previous comments from stakeholders have substantiated, the most important issues facing most mass transit public transport projects are serious cost overruns and fewer passengers than expected, which leads to insufficient fare revenues to break even or realize a small profit. Based on this information and other stakeholder inputs, the Board may wish to consider requesting the Methodology Panel to further improve the current draft revision of ACM0016. We thank you for your time and attention to this matter. Yours sincerely,

Matthew Woods Director

UNFCCC/CCNUCC CDM – Executive Board Page 4

Version 01/ 02 August 2011

Please use the space below to any mention any suggestions or information that you want to provide to the Board. In doing this please describe the exact reference source including the version (if any).

>> [replace this bracket with text, the field will expand automatically with size of text]

If necessary, list attached files containing relevant information (if any)

• Cities On The Move _A World Bank Urban Transport Strategy Review.pdf

Section below to be filled in by UNFCCC secretariat

Date when the form was received at UNFCCC secretariat

- - - - -

History of document

Version Date Nature of revision

01.1 09 August 2011 Editorial revision.

01 04 August 2011 Initial publication date.

Decision Class : Regulatory Document Type : Form Business Function : Governence

CITIES ON THE MOVE

A WORLD BANK

URBAN TRANSPORT

STRATEGY REVIEW

THE WORLD BANK

Washington, D.C.

© 2002 The International Bank for Reconstruction and Development / The World Bank1818 H Street, NWWashington, DC 20433Telephone 202-473-1000Internet www.worldbank.orgE-mail feedback@worldbank.org

All rights reserved. Published August 2002Manufactured in the United States of America1 2 3 4 05 04 03 02

The findings, interpretations, and conclusions expressed here are those of the author(s) and do notnecessarily reflect the views of the Board of Executive Directors of the World Bank or the govern-ments they represent.

The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries,colors, denominations, and other information shown on any map in this work do not imply on thepart of the World Bank any judgment of the legal status of any territory or the endorsement oracceptance of such boundaries.

Rights and Permissions

The material in this work is copyrighted. No part of this work may be reproduced or transmitted inany form or by any means, electronic or mechanical, including photocopying, recording, or inclusionin any information storage and retrieval system, without the prior written permission of the WorldBank. The World Bank encourages dissemination of its work and will normally grant permissionpromptly.

For permission to photocopy or reprint, please send a request with complete information to theCopyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to theOffice of the Publisher, World Bank, 1818 H Street, NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org.

ISBN 0-8213-5148-6

Cover photo: John Flora, 1995; Guangzhou, China.

Library of Congress Cataloging-in-Publication Data

World Bank.Cities on the move : a World Bank urban transport strategy review.

p. cm.“The author of the report is Ken Gwilliam”—Ackn.Includes bibliographical references.ISBN 0-8213-5148-6

1. Urban transportation. 2. Urbanization. I. Gwilliam, K. M. II. Title.

HE305. W67 2002388.4—dc21 2002025882

CONTENTS iii

CONTENTS

ACKNOWLEDGMENTS vii

ABBREVIATIONS AND ACRONYMS, CITIES AND COUNTRIES ix

EXECUTIVE SUMMARY xi

1 CONTEXT FOR THE REVIEW 1

2 URBAN TRANSPORT AND CITY DEVELOPMENT 5Urban Transport and City Efficiency 5Sources of Declining Urban Transport Performance 8Urban Transport in City Development Strategies 11Development Strategies for Making Urban Transport More Efficient 12Coordinating Sector Policies in City Development Strategies 21Conclusions: A Strategy for Urban Transport in City Development 22Notes 23

3 URBAN TRANSPORT AND POVERTY REDUCTION 25Urban Poverty and Social Exclusion 25Transport Patterns of the Urban Poor 25Pro-Poor Economic Growth and Poverty Reduction 28Focusing Infrastructure Policies 29Public Transport Service Planning for the Urban Poor 32Fare Policies, Subsidies, and Budget Constraints 34Conclusions: A Strategy for Poverty-Focused Urban Transport 36Notes 37

4 TRANSPORT AND THE URBAN ENVIRONMENT 39The Scale of the Issue 39Global Warming 40Noise and Other Disturbances 42Urban Air Pollution 42Conclusions: A Strategy for Urban Transport and Environment 60Notes 61

5 URBAN TRANSPORT SAFETY AND SECURITY 65The Scale of the Issue 65Safety 66Security 72Conclusions: A Strategy for Urban Transport Safety and Security 75Notes 76

iv CONTENTS

6 THE URBAN ROAD SYSTEM 77Elements of Road Strategy 77Road Maintenance 77Traffic Management 79Demand Management 83Infrastructure Provision 84Poverty Focus 88Conclusions: A Strategy for Roads 90Notes 91

7 PUBLIC ROAD PASSENGER TRANSPORT 93The Urban Bus Sector 93Conclusions: Toward a Strategy for Public Transport 106Notes 107

8 MASS RAPID TRANSIT 109The Scale of the Issue 109Objectives and Role of Mass Rapid Transit within the City Development Strategy 110Relationship to Urban Structure and Land Use 111Mass Rapid Transit and the Urban Environment 111Choice of Mass Rapid Transit Technology 112Ownership and Financing 117Public Transport Integration 121Pricing 121Conclusions: Toward a Strategy for Mass Rapid Transit 122Notes 123

9 THE ROLE OF NONMOTORIZED TRANSPORT 125The Importance of Nonmotorized Transport 125Walking Is Transport 126Walking on Wheels 126The Policy Package 128Municipal Institutions and Organization 132Conclusions: A Strategy for Nonmotorized Transport 134Notes 134

10 URBAN TRANSPORT PRICING AND FINANCE 135The Role of Prices in Urban Transport 135Charging for the Use of Infrastructure 135Public Transport Pricing and Finance 141Urban Transport Financing 146Conclusions: A Strategy for Urban Transport Pricing and Financing 150Notes 151

11 STRENGTHENING URBAN TRANSPORT INSTITUTIONS 153The Importance of Institutions 153Major Institutional Weaknesses 154

CONTENTS v

Organizational Options 156Capacity Building 165Conclusions: A Strategy for Institutional Reform in Urban Transport 167Notes 168

12 MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 169Development Challenges 169Strengthening the Focus on Poverty 170Facilitating Decentralization 171Mobilizing Private Participation 173Increasing Transport Safety and Security 175Protecting the Environment 176Conclusions 179Notes 182

APPENDIX: URBAN TRANSPORT PORTFOLIO IN THE WORLD BANK 183A.1 Urban Transport Projects under Preparation and Implementation as of

February 26, 2002 183A.2 Special Project Features 187A.3 Urban Portfolio: Project Cost and World Bank Group Loan and

Grant Amount Allocation 190

BIBLIOGRAPHY 193

INDEX 201

TABLES

2.1 A Categorization of City Circumstances 72.2 Estimates of External Costs of Road Transport 93.1 Poverty-Focused Urban Transport Interventions: Road Infrastructure 303.2 Income of Users of Rail Systems in Brazil 313.3 Poverty Impacts of Interventions: Rail Infrastructure and Management 323.4 Average Income of Users by Transport Mode 333.5 A Poverty-Focused Agenda: Public Transport Service Planning 353.6 Poverty-Focused Urban Transport Interventions: Finance Strategies 376.1 Functions and Responsibilities of a Typical Traffic Management Agency 826.2 Typical Functional Classification of Road and Path Networks 867.1 Efficiency of Bus Operations in Delhi, 1995 947.2 A Classification of Informal (Noncorporate) Urban Transport Operations 1028.1 Performance and Cost of Some Typical Mass Rapid Transit Systems 1138.2 Variation of Cost in Recent Metro Contracts 1178.3 Financial Performance of Some Metro Systems 118

11.1 Allocation of Strategic Functions 15411.2 Professional Organization for Municipal Transport Functions 158

vi CONTENTS

FIGURES

2.1 Motorization and Incomes: Growth of Cars versus Growth of per Capita Incomes in France, Japan, Spain, and the United Kingdom, 1950s to 1995; and the Relative Positions of Selected Other Countries, 1995 10

4.1 Overview of Ambient Air Quality in Selected Cities: Subjective Assessment of Monitoring Data, Various Years in the 1990s 44

7.1 Urban Public Transport Competition: Regimes and Transitions 977.2 Effects of Competition on Bus Transport in London 99

11.1 Effect of Transport Integration in Madrid 15711.2 A Typical Metropolitan Planning Process 161

BOXES

2.1 City Development Strategies 122.2 Matching Locational Characteristics and User Demands 172.3 Internal Consistency in Planning 182.4 The Market and Land Use and Transport Integration:

The Tama Garden City Development 204.1 Fuel Adulteration in Thailand 464.2 Electric Three-Wheelers in Kathmandu 494.3 The Benefits of Information: Lubricating Two-Strokes 534.4 Private Sector Vehicle Inspection and Maintenance in Mexico City 545.1 Road Accidents and the Poor in Sri Lanka 665.2 Saving Lives in Brasília 725.3 Crime, Violence, and Diminishing Mobility 736.1 Financing Urban Road Maintenance in the Kyrgyz Republic 786.2 Paving Roads in Low-Income Areas in Brazilian Cities 907.1 Introducing Competitively Tendered Franchises in Uzbekistan 1008.1 Leveraging Urban Transport Coordination in Brazil 1168.2 TransMilenio: Bogotá’s Bus Rapid Transit System 1198.3 The Hopewell Project in Bangkok 1209.1 Some Recent Experience with Facilities for Cycling in China 1289.2 The Bogotá Bicycle Master Plan 1299.3 Shova Lula (Ride Easy): A National Bicycle Transport Partnership 1309.4 Stakeholder Participation in Nonmotorized Transport Development in Tanzania 133

10.1 The Namsan Tunnels in Seoul: Simple Road Pricing Reduces Congestion and Finances Traffic Management 137

10.2 Electronics Improves Road-Pricing Efficiency in Singapore 13810.3 Recovering Fixed Costs: The Ramsey Pricing Rules 14311.1 Engaging Civil Society in Planning Resettlement of the Poor in Mumbai 165

This strategy review was undertaken on the initiative of John Flora, Director of the Transport andUrban Development Department of the World Bank. The author of the report is Ken Gwilliam,Economic Advisor to the Transport Department of the World Bank. The report has been discussedand approved by the Transport and Urban Sector Boards of the World Bank.

The work was undertaken in collaboration with the Urban Transport thematic group, which dis-cussed and agreed on the report in detail. Important contributors to this review process includedthe coordinators of the group, Ed Dotson and Richard Podolske, together with Patrick Bultynck,Sally Burningham, Robin Carruthers, Jean-Charles Crochet, Ben Eijbergen, Pierre Graftieaux, PaulGuitink, Jeff Gutman, Ajay Kumar, Zhi Liu, Gerhard Menckhoff, Slobodan Mitric, Hubert Nove-Josserand, Jorge Rebelo, Binyam Reja, Richard Scurfield, Graham Smith, Antti Talvitie, and LouThompson.

Colleagues from the Urban Development, Environment, and Energy sectors who made importantcontributions, particularly to chapters 2 and 4, include Samir El Daher, Fitz Ford, Vincent Gouarne,Sonia Hammam, Christine Kessides, Masami Kojima, Magda Lovei, and Margret Thalwitz.

In support of the review, several background reports were commissioned, funded by trust funds ordirect government commissioning. These reports, listed in the bibliography, are all separately avail-able on the World Bank Web site. The assistance of the governments of France, Germany, Japan,the Netherlands, Spain, and the United Kingdom is gratefully acknowledged.

A consultation draft of this report was discussed with major stakeholders in client countries, includ-ing governments, transport operators, and nongovernmental organizations, as well as with repre-sentatives of other international institutions in consultation meetings held at Accra, Ghana;Santiago, Chile; Budapest, Hungary; and Yokohama, Japan. Gerhard Menckhoff, Ben Eijbergen,Patrick Bultynck, and James Edezhath organized those meetings, with particular assistance from IanThompson of the Economic Commission for Latin America and the Caribbean (Santiago), staff ofthe Japanese Ministries of Transport and Communications (Yokohama), the KözlekedéstudományiIntézet Rt. Institute for Transport Sciences (Budapest), and the Sub-Saharan Africa TransportProgram (Accra). We are also grateful to the approximately 250 people from all over the world whowere involved in those consultations.

ACKNOWLEDGMENTS vii

ACKNOWLEDGMENTS

ALS Area licensing schemeCBD Central business districtCDS City development strategyCNG Compressed natural gas CO Carbon monoxideCO2 Carbon dioxideERP Electronic road pricingGDP Gross domestic productGEF Global Environment FacilityGHG Greenhouse gasGNP Gross national productI/M Inspection and maintenanceIBRD International Bank for Reconstruction

and DevelopmentIFC International Finance

CorporationITS Intelligent Transport SystemIU In-vehicle unit (Singapore)km/h Kilometers per hourLIL Learning and Innovation Loan LPG Liquefied petroleum gasLRT Light rapid transitmg/dl Micrograms per deciliterMMS Monthly minimum salary

MRT Mass rapid transitNDoT National department of transportation

(South Africa)NMT Nonmotorized transportNO Nitric oxideN2O Nitrous oxideNO2 Nitrogen dioxideNOx Oxides of nitrogenOECD Organisation for Economic

Co-operation and DevelopmentPM Particulate matterpphpd Passengers per hour in the

peak directionPRSP Poverty Reduction Strategy PaperRON Research octane numberSO2 Sulfur dioxideSOx Sulfur oxidesSRT Thailand State RailwaySSATP Sub-Saharan Africa Transport

ProgramUSAID U.S. Agency for International

Development WBI World Bank Institutewt ppm Parts per million by weight

ABBREVIATIONS AND ACRONYMS, CITIES AND COUNTRIES ix

ABBREVIATIONS AND ACRONYMS,CITIES AND COUNTRIES

ABBREVIATIONS AND ACRONYMS

CITIES AND COUNTRIES

Abidjan, Côte d’IvoireAccra, GhanaAddis Ababa, EthiopiaAhmedabad, IndiaAlmaty, KazakhstanBangalore, IndiaBangkok, ThailandBeijing, ChinaBeirut, LebanonBerlin, GermanyBilbao, SpainBishkek, Kyrgyz RepublicBogotá, Colombia

Brasília, BrazilBremen, GermanyBucharest, RomaniaBudapest, HungaryCairo, Arab Republic of EgyptCali, ColombiaCape Town, South AfricaCaracas, República Bolivariana de

VenezuelaCasablanca, MoroccoColombo, Sri LankaCopenhagen, DenmarkCotonou, Benin

x ABBREVIATIONS AND ACRONYMS, CITIES AND COUNTRIES

Curitiba, BrazilDakar, SenegalDamascus, Syrian Arab RepublicDar es Salaam, TanzaniaDhaka, Bangladesh Douala, CameroonGothenburg, SwedenGuangzhou, ChinaHanoi, VietmanHamburg, GermanyHarare, ZimbabweHong Kong, ChinaIbadan, NigeriaJakarta, IndonesiaJohannesburg, South AfricaKarachi, PakistanKarlsruhe, GermanyKathmandu, NepalKinshasa, Democratic Republic of CongoKolkata (formerly Calcutta), IndiaKraków, PolandKuala Lumpur, MalaysiaKunming, ChinaLa Paz, BoliviaLagos, NigeriaLeón, MexicoLeón, NicaraguaLiaoning, ChinaLima, PeruLomé, TogoLondon, United KingdomMadras (now Chennai), IndiaManila, Philippines Maputo, MozambiqueMaracaibo, República Bolivariana de

Venezuela

Medellín, ColombiaMexico City, MexicoMontevideo, UruguayMorogoro, TanzaniaMumbai (formerly Bombay), IndiaNairobi, KenyaNew Delhi, IndiaOuagadougou, Burkina FasoPorto Alegre, BrazilPhnom Penh, CambodiaPusan, Republic of KoreaQuito, EcuadorRecife, BrazilRio de Janeiro, BrazilRostov, Russian FederationRotterdam, NetherlandsSamarkand, UzbekistanSantiago, ChileSão Paolo, BrazilSeoul, Republic of KoreaShanghai, ChinaSingaporeStockholm, SwedenStuttgart, GermanySydney, AustraliaTehran, Islamic Republic of IranTunis, TunisiaUrumqi, ChinaVienna, AustriaVientiane, Lao People’s Democratic

RepublicWarsaw, PolandWuhan, ChinaYokohama, JapanZurich, Switzerland

A previous World Bank urban transport strategy paper concentrated on economic

and financial viability. “Urban Transport” (World Bank 1986) emphasized efficient

management of existing transport capacity, good traffic management, and

efficient pricing. It discouraged subsidies, recommended competition and

minimal regulation, and questioned the value to the urban poor of capital-

intensive projects that might not be cost effective in countries with limited

resources.

Subsequent sector strategy papers have takena broader view. The transport sector strategypaper “Sustainable Transport” (World Bank1996) emphasized the integrity of economic,social, and environmental dimensions of a sus-tainable transport policy. The urban develop-ment strategy paper “Cities in Transition”(World Bank 2000a) stressed that the livability ofcities depends on their being economicallycompetitive, financially sustainable, well gov-erned, and well managed.

This volume links the urban development andtransport sector strategies with a strongpoverty focus. Its objectives are (a) to develop abetter understanding of urban transport prob-lems in developing and transitional economies,(b) to articulate an urban transport strategyframework for national and city governments,and (c) to identify the role of the World Bank insupporting governments. It concentrates on theproblems of people who are very poor, not onlyin terms of income but also in terms of thebroader dimensions of social exclusion associat-ed with inaccessibility: inaccessibility to jobs,schools, health facilities, and social activities.

Some well-established urban trends continue.Urban population continues to expand at morethan 6 percent per year in many developing

countries. The number of megacities—citieswith over 10 million inhabitants—is expected todouble within a generation. More than one-halfof the developing world’s population, andbetween one-third and one-half of its poor, willthen live in cities. Per capita motor vehicle own-ership and use continue to grow by up to 15 to20 percent per year in some countries. Trafficcongestion and air pollution continue toincrease. Pedestrian and other nonmotorizedtransport (NMT) continue to be poorly served.Increased use of private vehicles has resulted infalling demand for public transport and a con-sequent decline in service levels. Sprawlingcities are making the journey to work excessive-ly long and costly for some of the very poor.

The context has changed in some significantrespects since 1986. Cities are increasinglyinvolved in trading patterns on a global scale,which makes the efficiency of their transport sys-tems more critical. At the same time, responsi-bility for urban transport is being decentralizedto the cities, which are often strapped for cashand are institutionally ill prepared for the newchallenges. Under these conditions the financialstate of public transport has deteriorated drasti-cally in many countries. The safety and securityof urban travelers are emerging problems, par-ticularly in Latin America.

EXECUTIVE SUMMARY xi

EXECUTIVE SUMMARY

xii EXECUTIVE SUMMARY

THE FUNDAMENTAL PARADOX OFURBAN TRANSPORT STRATEGY

Urban transport can contribute to povertyreduction both indirectly, through its impacton the city economy and hence on economicgrowth, and directly, through its impact on thedaily needs of poor people. However, urbantransport exhibits a fundamental paradox. Howcan a sector with such an obvious excess ofdemand over supply and with such a heavyinvolvement of private suppliers of service fail socompletely to meet the aspirations of bothpoliticians and citizens? Why has it not beenpossible to mobilize commercial initiative toyield the kind of revolution in service quality andcost that has been achieved in the telecommu-nications, water, and energy sectors? Finally, whydoes increasing affluence seem to have theeffect of reducing the quality of travel, at leastfor poor people?

Urban growth increases transport costs. Fromthe viewpoint of efficiency and growth, it is nottoo difficult to characterize the central problem.Economies of agglomeration generate thegrowth of cities. As cities grow and becomericher, vehicle ownership and use grow morerapidly than the available road space, resultingin increased congestion and traffic-generatedair pollution.

Urban growth often has perverse distribution-al effects. As cities expand, the price of moreaccessible land increases. Poor people areforced to live on less-expensive land, either ininner-city slums or on city peripheries. As aver-age incomes grow and car ownership increas-es, the patronage, financial viability, and even-tually quality and quantity of public transportdiminishes. Motorization, which is permittedby the growth process, may thus also makesome poor people even poorer. In particular, inthe absence of efficient congestion pricing forroad use, piecemeal investment to eliminatebottlenecks will almost certainly benefit the rel-atively wealthy at the expense of the poor.

An eclectic strategy is proposed. The strategyincludes four main ways to address these prob-lems: (a) structural change, (b) improved opera-tional efficiency of the transport modes, (c) bet-ter focusing of interventions to assist the poor,and (d) policy and institutional reform.

STRUCTURAL CHANGE

Deconcentration has a limited role to play. Themost fundamental structural response is to tryto shift activity away from megacities, concen-trating new development in medium-size cities.Unfortunately, it is not clear at what city size theeconomies of agglomeration run out or how apolicy of deconcentration can be effectivelyimplemented. Nevertheless, central govern-ments can encourage the development ofsmaller regional hubs by eliminating fiscal andpublic expenditure distortions, including elimi-nation of price distortions in land and transportmarkets, such as the underpricing of congestedroad space and the absence of full-cost con-nection charges and impact fees for land devel-opment. They can also lead by the location oftheir own activities.

Improved structure within cities can con-tribute greatly. A less-radical approach empha-sizes coordination of land use and transportinfrastructure and service planning, to ensureprovision of adequate and well-structured roadspace as the city grows. This requires improveddevelopment control skills and practices at thecity level. Critics of this approach argue thatsuch an emphasis on road capacity fosters alevel of motorization that will create depend-ence on the automobile, and will eventuallyovertake space availability. In any case, it isunlikely to be socially or environmentallyacceptable to balance supply and demandsolely by increasing road capacity in largercities.

Good road infrastructure does not necessarilymean total auto dependence. Indeed, it is the

EXECUTIVE SUMMARY xiii

combination of land-use and transport planningthat has made it possible for some cities to rec-oncile high mobility with high quality of urbanlife. In order to achieve that reconciliation, traf-fic has been restrained (as in Singapore, by roadpricing) and has been managed to maintainsafe, efficient, and environmentally acceptablemovement of people, not just of vehicles. Thisimplies prioritization of infrastructure to protectmovements of public transport and NMTagainst unrestricted expansion of private motor-ized trips (as in Bogotá, Colombia, and Curitiba,Brazil, through busway systems). In these more-constrained circumstances, rigorous appraisalof investments in road capacity needs to takeinto account (a) the effects of induced traffic onbenefits; (b) the benefits to, and disbenefits of,NMT; and (c) the environmental impacts.

IMPROVING THE OPERATIONALEFFICIENCY OF TRANSPORT

To improve the efficiency of transport, theneeds of each mode must be addressed—theroad system, NMT, public passenger transport,and mass transit. In addition, the role of the pri-vate sector as a means of promoting efficiencydeserves special attention.

THE ROAD SYSTEMEven in highly congested cities, urban roadtransport efficiency can be improved throughbetter system management. Although rapiddevelopment of technology has reduced thecost—as well as the maintenance and opera-tional skill requirements—of modern trafficmanagement techniques, many cities are stilltoo poorly organized and inadequately staffedto make effective use of this development.Both technical assistance and investment arecapable of yielding high returns in this field, aslong as fundamental institutional and humanresource problems are addressed.

Urban road decay is a serious problem in manycountries. Road decay contributes to conges-

tion and increasing operating costs. It oftenarises from jurisdictional conflicts—such as con-flicts over which authority is responsible forwhich roads, lack of clear ownership of neigh-borhood roads, or inadequate allocation forurban roads from the national road fundsthrough which road funding is channeled.

NONMOTORIZED TRANSPORTNMT is systematically underrecognized.Walking still accounts for the largest proportionof trips taken, although not of distance trav-eled, in most low- and middle-income coun-tries. All income groups are involved. Despitethis fact the welfare of pedestrians, and partic-ularly the welfare of mobility-impaired pedestri-ans, is frequently sacrificed in planning toincrease the speed of the flow of vehicles.Cycling is similarly disadvantaged. Without acontinuous network of secure infrastructure,people will not risk bicycle travel. Withoutusers, investment in infrastructure for cyclingmay appear wasteful.

A comprehensive vision and action plan forNMT is required. In the planning and manage-ment of infrastructure, the excessive emphasison motorized transport may be redressed by (a)clear provision for the rights as well as respon-sibilities of pedestrians and bicyclists in trafficlaw; (b) formulation of a national strategy forNMT as a facilitating framework for local plans;(c) explicit formulation of a local plan for NMTas part of the planning procedures of municipalauthorities; (d) provision of separate infrastruc-ture where appropriate (such as for safe move-ment and secure parking of vehicles); and (e)incorporation of standards of provision for bicy-clists and pedestrians in new road infrastructuredesign. Incorporation of responsibilities forprovision for NMT should also be included inroad fund statutes and procedures.

Traffic management should be focused onimproving the movement of people rather thanon improving the movement of motorizedvehicles. In order to achieve that goal, police

xiv EXECUTIVE SUMMARY

need to be trained to enforce the rights of NMTin traffic priorities as well as in recording andpreventing accidents. Furthermore, the devel-opment in poor countries of small-scale creditmechanisms to finance bicycles, credit mecha-nisms that are increasingly successful in ruralareas, might also be developed in urban areas.

PUBLIC PASSENGER TRANSPORTPublic transport is for all. Concentrating on thetransport modes of poor people in middle-income countries essentially means the provi-sion of affordable forms of public transport,both formal and informal. But it should not beviewed as only for the poor, as the importanceof public transport to all income groups inmany rich European cities demonstrates.Improving efficiency in public transport must beconcerned not only with keeping costs downbut also with providing a flexible frameworkwithin which the less poor as well as the verypoor can use public transport with confidenceand comfort.

Most urban public transport is road based.Bus lanes and automatic priority at intersec-tions can improve public transport perform-ance significantly, but these solutions tend tosuffer from inadequate enforcement by police,who are untrained in traffic planning and man-agement. In contrast, exclusive busways indeveloping countries have proved to be capa-ble, except in very high traffic volume corridors,of performance nearly equivalent to rail-basedsystems but at much lower cost.

Pricing and financing issues are at the heart ofpublic transport problems. Formal bus opera-tions face financial collapse in many countries,partly as an unintended consequence of good-hearted but wrong-headed fare and servicecontrols. Some prescriptions can easily bemade to forestall this. General fare controlsshould be determined as part of a comprehen-sive city transport financing plan, and theireffect on the expected quality and quantity ofservice carefully considered. Fare reductions or

exemptions should be financed on the budgetof the relevant line agency responsible for thecategories (health, social sector, education,interior, and so on) of the affected person.Modally integrated fare schemes should beassessed for their impacts on poor people. It isin the interests of poor people for sustainablefinancing and efficient targeting of public trans-port subsidies to be paramount.

There is a rich agenda of urban public trans-port policies that is both pro-growth and pro-poor. The recent decline in both the quality andquantity of public transport has resulted partlyfrom the absence or disappearance of a securefiscal basis for support. Public transport, how-ever, can be improved in many ways that areconsistent with the fiscal capabilities of eventhe poorest countries. Giving priority to publictransport in the use of road space makes publictransport faster and more financially viable.

Competition is pro-poor. Supply costs can bereduced through competition between privatesector suppliers. In Buenos Aires the urban railsystem has been revolutionized through con-cessioning. Regulated competition in the busmarket has also worked well in cities such asBuenos Aires and Santiago—but care is need-ed in system design. Total deregulation in Lima,although it has increased supply, has worsenedroad congestion, the urban environment, anduser safety and security. The lesson is that it isnot privatization or deregulation per se thatimproves public transport, but rather the intro-duction of carefully managed competition, inwhich the role of the public sector as regulatorcomplements that of the private sector as serv-ice supplier.

Cities should strive to mobilize the potentialof the informal sector. Informally suppliedsmall vehicle paratransit (publicly available pas-senger transport service that is outside the tra-ditional public transport regulatory system) isoften dominant in providing for dispersed trippatterns and in flexibly addressing the

EXECUTIVE SUMMARY xv

demands of poor people, particularly in low-income countries, but it is typically viewed aspart of the problem of public transport and notpart of the solution. Certainly, anticompetitiveor antisocial behavior should be controlledthrough quality controls and enforcement, butits potential can be better mobilized throughlegalizing associations and through structuringfranchising arrangements to give the smalloperator an opportunity to participate in com-petitive processes.

MASS TRANSITRail-based mass transit systems have a role toplay in very large cities. Rail-based mass transitsystems are less congesting than are road-based systems and can be very important forthose who are peripherally located and havelong journeys to access employment in thecities. In Latin America, in particular, rail-basedsystems carry significant numbers of very poorpeople. The Bank has financed several majorurban rail developments in the past decade,typically in metros and existing suburban rail-way refurbishment but occasionally in new con-struction. Often the restructuring of bus servic-es, which eliminates direct competition and canharm the interests of poor bus-users unlessskillfully planned, supports the rail-based sys-tems. The position that has been adopted isthat such developments must be integratedinto a comprehensive urban transport strategyand that arrangements should include physicaland fare integration between modes, to ensurethat the poor are not excluded from or disad-vantaged by the Bank’s investments.

Urban rail-based systems should be cautiouslyappraised. Urban rail-based systems are costlyto build and operate, are more expensive forthe passenger to use than road-based modes,and can impose a large burden on the citybudget. It remains appropriate, therefore, toadvise cautious examination of the fiscal sus-tainability of rail investments and their impacton poor people before making expensive com-mitments. The most critical lesson the Bank has

learned is that mass transit investment deci-sions should be driven by a thorough examina-tion of strategic objectives of technologicalalternatives, and financial implications, and notby short-term political or commercial oppor-tunism.

THE ROLE OF THE PRIVATE SECTOR Private financing of urban transport infrastruc-ture is possible. Recognizing the burden ofinvestments in major roads and metros onmunicipal budgets, cities such as Bangkok,Buenos Aires, and Kuala Lumpur have alreadymanaged to secure private capital finance forthem. Experience so far has shown that thisrequires very high demand for faster movementin the affected corridor and a realistic stance bygovernment on the relationship between pricecontrols and commercial profitability.Experience has also shown opportunistic devel-opment on an ad hoc basis to be damaging,and usually costly to the public purse. Masstransit systems, in particular, appear to yieldgreatest benefit when they are incorporatedinto a citywide price-level and structure plan inwhich the full cost of new mass transit invest-ments on the municipal budgets, on fares, andon poor people has been estimated inadvance.

Planning and regulatory arrangements for pri-vate participation in urban transport are fun-damental. The interaction of transport withland use requires its careful integration into theplanning of metropolitan structure and financewithin a comprehensive long-term plan for thecity. The public sector must set a strategy; iden-tify infrastructure projects and describe them insome detail; and confirm the acceptability ofenvironmental consequences, tariffs, and anycontingent changes to the existing transportsystem. It must acquire the necessary land andrights-of-way, ensure development permis-sions, commit funding, and provide some nec-essary guarantees. Physical coordination (toachieve convenient modal interchange) andfares coordination (to keep public transport

xvi EXECUTIVE SUMMARY

attractive and to protect poor people) need tobe embodied in a comprehensive transportstrategy plan that recognizes the relationshipsbetween modes of transport.

BETTER FOCUSING OFINTERVENTIONS TO ASSIST THEPOOR

There are two possible approaches whendesigning poverty-targeted transport interven-tions—directly serving the locations wherepoor people live and work, and targeting dis-advantaged groups. In addition, institutionsmust address two issues that have a particularimpact on the poor—the polluted urban envi-ronment, and safety and security.

SERVING THE LOCATIONS WHEREPOOR PEOPLE LIVE AND WORKTransport improvements can be focused onwhere poor people live and work. Theseimprovements may involve concentrated effortsto improve access to slum areas or to improvepublic transport to peripheral locations. TheBank-supported Pavement Program in Low-Income Areas (Programa de Pavimentacao deBaixo Custo em Areas de Baixa Renda—PROPAV) in Brazil proved highly successful, andwas extended throughout the country, as well asto other Latin American countries.

Leakage through land rent changes must betaken into account. Transport investments orservice improvements change the structure ofland values. If there is strong competition forthe use of land and highly concentrated own-ership of land, rents increase in improvedareas and the benefits of transport improve-ments accrue to rich landowners rather than topoor land occupants. Some investments—such as improvements in bus or NMT sys-tems—are less likely to drive poor people outto more distant, less-expensive locations thanare others—such as primary roads or morehighly priced, mass transit systems. This find-

ing further emphasizes the need for transportto be part of a comprehensive urban develop-ment strategy.

TARGETING DISADVANTAGED GROUPSTransport provision can be part of a socialsafety net. A complementary approach is tofocus on the specific categories of disadvan-taged people. Given the overwhelming impor-tance of the ability to access employment, thework journeys of poor people may be a primetarget for support. The cost of ensuring thatthese trips are affordable may be shifted to theemployer (as with the “vale-transporte” inBrazil) or the state (as with the commuter sub-sidy system of South Africa). Although they maybe less-than-perfectly targeted (for example,the vale-transporte misses very poor informalworkers), may distort residential location incen-tives, and are inferior to direct income transfers,targeted transport subsidy arrangements maybe the best practicable safety net for poorworkers.

Low income is not the only form of depriva-tion. Gender confers some particular disadvan-tages in terms of diffused trip patterns and tim-ings, as well as particular vulnerability to safetyand security problems. Age and infirmity poserather different problems, calling for sensitive“inclusive” design of physical facilities. Bothlocational resettlement and occupational rede-ployment impinge in a particularly harsh way onpoor people, requiring adequate safety nets.

Fare controls can do more harm than good.Experience teaches two important lessonsabout what not to do with respect to fare con-trols. First, controlling fares in the absence ofrealistic analysis of, and provision for, theresource needs of that social strategy actuallydestroys public transport service and maycause serious harm to some poor people.Second, cross-subsidy within public sectormonopolies does not eliminate the fundamen-tal resource problem, and instead adds someextra burden of inefficiencies in supply.

EXECUTIVE SUMMARY xvii

POVERTY OF “LIFE QUALITY”: TRANS-PORT AND THE URBAN ENVIRONMENTPoor people tend also to be the most vulnera-ble to environmental pollution. The most dam-aging pollutants are lead, small suspended par-ticulate matter, and in some cities, ozone. Localair pollution from transport in developing coun-tries contributes to the premature deaths ofover 500,000 people per year, and imposes aneconomic cost of up to 2 percent of grossdomestic project (GDP) in many countries. Astrategy for improvement of the effects thaturban transport has on the environment is thusnot a luxury to be afforded at the expense ofpoor people, but an important element of anurban transport strategy. The Intergovern-mental Panel on Climate Change (IPCC) alsoforecast that developing countries will sufferdisproportionate costs of from 5 to 9 percent oftheir GDP should the global level of carbondioxide double (IPCC 1996).

Understanding of the environmental impactsof urban transport remains deficient. There aresome clear technological priorities. While it isgenerally preferable to concentrate on perform-ance standards, rather than on specific technol-ogy preference, there are also some relativelyclear technological priorities for the sector.These include the elimination of lead from gaso-line, the replacement of two-stroke motorcycleswith four-stroke motorcycles, and the eliminationor cleaning up of high-mileage, heavily pollutingvehicles. The Bank can help with technical assis-tance in these fields and, in some cases, with thefinancing of infrastructure and incentive mecha-nisms to stimulate change.

There is no quick technological fix for devel-oping countries. Local air quality can beimproved in the long run by new fuel and vehi-cle technologies. In the short run, however, thevehicle stock is dominated by an older genera-tion of technology, which is often badly main-tained. In some countries the emphasis onidentifying and acting to improve the worst,highest-mileage polluters—often buses, taxis,

and some trucks—has helped. Inspection andmaintenance programs, if undertaken by tech-nologically efficient instruments in a corruption-free context, can have great impacts. At theextreme there are assisted, or forced, scrap-page schemes.

Some robust “win-win” environmental strate-gies exist for the urban transport sector. Goodtraffic management can reduce environmentalimpact as well as congestion. Tax structurereform can encourage the use of cleaner fuelsand stimulate better vehicle maintenance. Thisreform, however, requires the design of fiscalmeasures to handle problems associated withthe use of fuels (for example, kerosene, which isused in several sectors), and to handle the asso-ciated conflicting policy objectives, such asthose associated with the taxation of diesel fuel(see the more detailed discussion in the maintext of the report). The integration of transportinterventions in general municipal develop-ment packages may offer better leverage in thisrespect than the integration of transport-specif-ic projects.

SAFETY AND SECURITYRoad accidents are a global pandemic. Nearly0.5 million people die and up to 15 million peo-ple are injured in urban road accidents indeveloping countries each year, at a direct eco-nomic cost of between 1 and 2 percent of GDPin many countries. Accidents occur widely onroads between intersections rather than beingconcentrated at intersections, as is the case inindustrialized countries, and the majority of vic-tims are poor pedestrians and bicyclists.

Adequate data are the basis for policy formu-lation and implementation. The first steps toimprove traffic safety are the development ofnational road accident data collection and analy-sis capability, and the formation of institutionalarrangements to ensure that the data are trans-mitted to those who need them for policy pur-poses. Accident frequency and severity can bereduced by improved road design and traffic-

xviii EXECUTIVE SUMMARY

management policies. While some infrastructureinvestment is specifically safety oriented (such asinfrastructure for NMT in Lima, or grade-separat-ed railway crossings in Buenos Aires), there is astrong case for mandatory safety audits in thedesign process for all transport infrastructure.Improved medical response can be achieved bysome relatively inexpensive and simple institu-tional innovations. Increasing safety awarenessto change traffic patterns and pedestrianbehavior requires development and training ofstaff for specific road-safety coordinating agen-cies or councils, at both the national and munic-ipal levels.

Personal security is a growing social problemin many countries. While this problem encom-passes much more than the transport sector, it isimportant to analyze the nature and significanceof insecurity in the urban transport sector and todevise policy instruments to counter it. Thatmight include collection and analysis of data onpersonal security in the transport sector toenhance official awareness of the problem, andmight include commitment of police authoritiesto arrest and the courts to appropriately penalizeoffenders. Strengthening public participation inprojects, particularly at the neighborhood level,is important. Some transport policy initiatives cancontribute directly to better personal security. Forexample, street lighting—designed to improvepedestrian security—can be included in streetimprovement, and particularly in slum-upgrad-ing, projects. Franchise conditions for publictransport can give incentives for improved atten-tion to security by public transport operators.

POLICY AND INSTITUTIONALREFORMS

Technical measures alone are unlikely to resolvethe fundamental paradox of a sector’s combin-ing excess demand with inadequately financedsupply. Improvements in the efficiency of roads,vehicles, public transport operations, and trafficmanagement can undoubtedly improve the

efficiency of urban transport. This will not beenough, however, because of three structuralcharacteristics that distinguish urban transportfrom most other urban service sectors. Thesecharacteristics are (a) the separation of infra-structure from operations, (b) the separation ofinteracting modes of transport, and (c) the sep-aration of infrastructure finance from infrastruc-ture pricing. What is required, therefore, is anintegrated package of strategies for infrastruc-ture pricing, service pricing, and urban trans-port system financing, founded in well-designed institutions within an appropriatepolitical framework.

SEPARATION OF INFRASTRUCTUREFROM OPERATIONSCharging for road infrastructure is the core ofa strategy for both efficient allocation ofresources and sustainable finance. Congestionincreases private transport costs and con-tributes to the decline of public transport serv-ice. While these two phenomena are logicallyconnected, in most cities they are institutionallyand financially separated. In principle, vehicularusers of congested urban road space should becharged a price at least equal to the short-runmarginal cost of use, including congestion, roadwear and tear, and environmental impacts.

In the absence of direct charging, fuel taxationshould be structured concurrent with vehiclelicense duties to give a proxy charge for roaduse and its external impacts. In practice, arange of direct and indirect mechanisms is usedto charge for road use. The most common ofthese mechanisms—the fuel tax—reflects glob-al warming impacts well, but is a poor surrogatefor either congestion or road-maintenanceimpact pricing. Nevertheless, if it is the bestproxy there is, the fuel tax should be structuredto reflect its relative contributions to urban airpollution, again in conjunction with the struc-turing of vehicle license duties.

Parking charges should be related to an over-all infrastructure pricing strategy. Although

EXECUTIVE SUMMARY xix

they are also a poor proxy for congestioncharges, parking charges should, in any case,always cover the full opportunity-cost of landused for parking. Where parking policy is theonly available proxy for congestion, pricingcontrols need to cover all forms of parkingspace (including that provided privately byemployers for their employees).

Direct charging for roads requires carefulpolitical and administrative preparation.Although cordon pricing and tolling of specificroads is a step in the right direction, the long-term solution lies in more systematic conges-tion charges. Of course, it is not easy to raiseprices or taxes, particularly for goods that havetraditionally been viewed as free. For instance,resistance to increased fuel prices in theRepública Bolivariana de Venezuela in the late-1980s was very violent. Riots following anincrease in public transport fares in Guatemalain 2000 cost five lives. This suggests that suchincreases in charges must be linked with a per-ceptible improvement in provision of services.There would remain a large education require-ment to explain the link between the increasedcost and the improvement of services, and tooffer realistic choices of alternatives. The sec-ond part of the integrated solution thus refersto service provision and pricing.

SERVICE PROVISION AND PRICING Pricing principles for public transport modesshould be determined within an integratedurban strategy. This means that they shouldreflect the extent to which road infrastructure isadequately charged. Given the high level ofinteraction among modes, and the prevalentundercharging of road use, financial transfersbetween roads and public transport services—and between modes of public transport—arepotentially consistent with optimal pricingstrategy.

Subsidies or compensation payments do notmean that there should be a monopoly suppli-er of transport services. In the interests of effi-

cient service supply, transport operators shouldoperate competitively, with purely commercialobjectives, and with financial transfers achievedthrough contracts between municipal authoritiesand operators for the supply of services. Anynoncommercial objectives imposed on operatorsshould be compensated directly and transpar-ently, where appropriate, by the nontransportline agencies in whose interests they areimposed. Above all, in the absence of appropri-ate contracting or other support mechanisms,the sustainability of public transport serviceshould be paramount, and should generally haveprecedence over traditional price regulationarrangements. The completion of an integratedpolicy thus requires an integrated urban trans-port financing system.

URBAN TRANSPORT SYSTEM FINANCINGUrban transport financing should be fungible.Given the interaction among modes, there is astrong case for treating the urban transport sys-tem as an integrated whole. Because neithercongestion nor environmental impacts are cur-rently subject to direct charges in many coun-tries, optimizing the performance of the sectoras a whole might justify using revenues raisedfrom private automobile users to fund improve-ments in public transport. Private sector financ-ing for transport infrastructure, raised throughcompetitive tendering of concessions, may besupported by public contributions as long asthese have been subject to proper cost-benefitanalysis.

There are different ways of securing fungibili-ty of funding. In a well-managed unitaryauthority, such as in Singapore, this occursthrough the normal budgetary process. In morecomplex, multitiered administrative systems,achieving this flexibility may require the pool-ing of urban transport financial resources with-in an urban transport fund administered by astrategic transport authority at the municipal ormetropolitan level. Under such an organization,all local transport-user charges, including con-

xx EXECUTIVE SUMMARY

gestion charges and any allocations of localtaxes or intergovernmental transfers for trans-port, should normally be made to the fund.

Urban transport funds do not imply earmark-ing of taxes. Earmarked taxes, such as the pay-roll tax on employers, that supports the publictransport agency Régie Autonome desTransport Parisiens (RATP) in Paris, have theadvantage of a secure legal and budgetaryfoundation, and are often the basis on whichsound long-term service planning can beundertaken. However, the value of having anintegrated urban transport fund does notdepend on any specific tax source being ear-marked for transport. Moreover, in order todevelop the credibility of the fund, and particu-larly to gain political and popular support forthe payment of congestion charges, it is essen-tial that the objectives and scope of an urbantransport fund be clearly defined, that alloca-tions be subject to rigorous appraisal, and thatthe operations of the fund be transparent.

INSTITUTIONSPolicy integration has significant institutionalimplications. In the interests of urban transportintegration and sustainability, developing coun-tries could therefore profitably move towardprices reflecting full social costs for all modes, toa targeted approach to subsidization reflectingstrategic objectives, and to an integration ofurban transport funding, while still retainingsupply arrangements for individual modes thatgive an important incentive to operational effi-ciency and cost-effectiveness. The implementa-tion of such a policy package has significantinstitutional implications, requiring close coordi-nation both between jurisdictions and betweenfunctions, as well as between private and publicsector planning and operating agencies.

The basis for institutional coordination is oftenvery weak. Few cities have a strategic agencyfor land-use and transportation planning, or acompetent traffic management unit. Trafficpolice are therefore often involved with traffic

management planning, for which they are illequipped and untrained. Public transport plan-ning and regulation is also often tied to opera-tions. The few institutions that do exist tend tobe understaffed and their staff poorly trained.

Urban transport institutions need both restruc-turing and strengthening. Action is required ontwo levels. First, authorities need to recognizewhat kind of technical organization is necessaryto address urban transport issues. Second, theorganizations need adequate human, as well asphysical, resources to perform their functions.While no single institutional blueprint for publictransport is appropriate for all countries, there isenough experience to establish some generalprinciples for the reduction of institutionalimpediments to effective policy integration.

Jurisdictional coordination may be facilitatedthrough the clear establishment in law of theallocation of responsibility between levels ofgovernment. Formal institutional arrangementscan be made for collaboration where multiplemunicipalities exist within a continuous conur-bation. The process of decentralization indeveloping countries may offer an excellentopportunity to address the problems. In partic-ular, intergovernmental transfers need to becarefully planned to be consistent with the allo-cation of responsibility, but structured to avoiddistorting local priority setting. Central govern-ments might also encourage coordination atthe metropolitan level; in France, for example,the central government made both local taxa-tion powers and intergovernmental transfersconditional on appropriate jurisdictional andfunctional collaboration.

Functional coordination should be based on astrategic land-use and transport plan. Detailedplanning, both of transport and land use, shouldbe aligned with a municipal or metropolitanstructure plan. Coordinated operation is furtherenhanced by the clear allocation of functionsamong agencies, with the more strategic func-tions being retained at the metropolitan level.

EXECUTIVE SUMMARY xxi

Obligations statutorily imposed on local author-ities should be linked to specific channels offinance (such as direct line agency funding ofreduced public transport fares). Responsibilityfor traffic safety should also be explicitly allocat-ed, with an institutional responsibility at thehighest level of the local administration. Trafficpolice should be trained in traffic managementand safety administration, and involved in trans-port and safety policy planning.

Responsibility for planning and operating pub-lic transport should be institutionally separat-ed. For effective involvement of the private sec-tor, technical regulation should be separatedfrom procurement and economic regulation. Aclear legal framework should be established forcompetition in public transport supply, either inthe market or for the market. Operations shouldbe fully commercialized or privatized, and thedevelopment of new competitive private suppli-ers of service encouraged through legal recog-nition of associations, and so on. The public sec-tor should develop strong service procurementand contract enforcement skills.

POLITICS, PARTICIPATION, AND PERFORMANCEDecentralized democratic process must becomplemented by high technical competence.Ultimately, transport policy formulation involvesan element of tradeoff between conflictinginterests. It is therefore bound to be a politicalprocess. Too often (not least in Latin America)bad investments have been made, and seriousurban transport issues trivialized, by the politi-cal process. Cities that have exhibited goodtransport planning and management, such asCuritiba and Singapore, have often developedunder strong leadership and have been found-ed on a high level of technical and professionalcompetence in the planning function. Thequestion is how to reconcile coherent technicalvision with more decentralized and fragmenteddemocratic processes.

Public participation and technically strongplanning can be complementary. The develop-ment of public participation and consultation,in parallel to the local democratic process, is animportant means of improving local policydesign. This may occur through advance expo-sure of plans to a free press and other media, aswell as through more formal processes of pub-lic consultation or public inquiry. For small-scale, very localized, infrastructure projects itmay be possible to incorporate local prefer-ences in the design process itself. Public trans-port users may also be involved in service fran-chising arrangements by complaints and con-sultation processes and by linking bonus pay-ments for franchised operators to public ormedia appraisal. At a more strategic level, andfor larger, more complex projects, consultationoften functions more as a means of trying toreconcile inherently competing or conflictinginterests; it is nevertheless central to the devel-opment of consensus-based city developmentstrategies.

Public participation must be timely and wellstructured. Developing strategic involvementrequires action at two levels. First, the publicprocesses must be organized to facilitate time-ly but well-informed consultation. Second, par-ticularly where formal local political processesare weak, the existence of effective local com-munity groups is extremely important. In devel-oping countries, such groups are often welldeveloped in rural areas but much less so incities. As both policy and financial responsibili-ty for urban development is decentralized tothe cities, it is thus possible to create institu-tional and financial arrangements that betterreflect the complex interactions both within theurban transport sector and between urbantransport and the rest of urban developmentstrategy. It is only on such a carefully consid-ered institutional and financial basis that thefundamental paradox of urban transport can beresolved.

CONTEXT FOR THE REVIEW 1

CONTEXT FOR THE REVIEW1

The primary focus of all World Bank activity is thereduction of poverty. The essence of its approach,embodied in the comprehensive developmentframework and in the poverty reduction strate-gies being developed with the highly indebtedpoor countries, is a holistic view of the develop-ment process, recognizing the interdependenceamong sectors, concentrating on the weak links inany particular country, and aiming to better coor-dinate the activities of the many agencies involved.

Economically, transport is the lifeblood of cities;in most countries, including developing coun-tries, cities are the major sources of the nationaleconomic growth. Cities, and the growth of cities,make poverty reduction possible. Poor transportinhibits the growth of cities. Socially, transport isthe means of (and the lack of transport is theimpediment to) accessibility to the jobs, health,education, and social services that are essentialto the welfare of the poor; inaccessibility emergesas a major cause of social exclusion in studies ofthe poor in urban areas. Urban transport strat-egy can thus contribute to poverty reduction boththrough its impact on the city economy, andhence on economic growth, and through its directimpact on the daily needs of the very poor.

Urban transport is already an important part ofthe Bank portfolio. At the end of 2000 there were

48 active projects with urban transport compo-nents and an additional 12 projects at a suffi-ciently advanced stage of preparation for WorldBank investment to be estimated. One-half ofthe 48 active projects were registered as urbantransport projects. The total sum of Bank lend-ing in these projects is estimated at $4.4 billion.1

Of that amount 35.3 percent was accounted forby metro and suburban rail investments (most ofthese in Latin America), and 13.7 percent byinvestments in buses, busways, and other high-occupancy vehicle facilities. An additional 19.2percent of the investment was on new road con-struction (most of this investment in China), and15.4 percent on road and bridge rehabilitationand maintenance. In addition, the InternationalFinance Corporation has, since 1995, invested inseven urban transport projects, including tollroads, buses, and a metro project, as well as anumber of major airport and seaport develop-ments in capital cities.

The context for this review is the progressivedecentralization of political and financial respon-sibility for urban development, either by delib-erate institutional shifts of decentralizationpolicy—for example, in Brazil—or by political andeconomic reality—for example, in many coun-tries of the former Soviet Union. As a recent WorldBank urban development strategy paper argues

Cities in the developing world are growing rapidly; with this growth comes

increased congestion. At the same time, public transport is declining in many

cities to the detriment of the city economy, its environment, and the welfare of

its poorer inhabitants. The purposes of this review are (a) to refocus urban

transport strategy on the issues of urban poverty, and (b) to offer a vehicle for

enhancing collaboration between the World Bank and other agencies involved in

urban development lending or aid.

2 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

(World Bank 2000a), decentralization willinevitably force city governments to confront theissues of fiscal balance and creditworthiness(“bankability”) and of good governance of theirassets, as a necessary condition for improvingthe competitive position of the city’s economyand the social and environmental conditions (“liv-ability”) for its inhabitants. Fostering the devel-opment of these institutional characteristics isthe primary thrust of the urban developmentstrategy paper (World Bank 2000a). That strategypaper did not, however, attempt to translate thephilosophy into specific service supply strategiesfor different sectors. That is what this volumeattempts to do: to view urban transport in thewider context of city development and within acomprehensive development framework that putsincreased emphasis on poverty and on enhancedcollaboration between the World Bank and othermultilateral and bilateral agencies involved indevelopment lending or aid.

The last World Bank urban transport policy paper(World Bank 1986) emphasized the importanceof planning and managing infrastructure and traf-fic to secure economically efficient urban move-ment. Since then, a broader perspective ontransport sector policy has been developing. Inlate 1996 the Bank published its general trans-port policy paper, which emphasized the essen-tial integrity of economic, social, and envi-ronmental dimensions of a sustainable transportpolicy (World Bank 1996).

The objectives of this urban transport reviewvolume are therefore (a) to develop a bettercommon understanding of the nature and mag-nitude of urban transport problems, particularlywith respect to the poor, in developing and tran-sitional economies, and (b) to articulate a strategyto assist national and city governments to addressurban transport problems within which the roleof the World Bank (and other agencies) can beidentified.

This is not a narrow technical matter. There issome tension among the objectives of growth

(of which motorization is, to some extent, anenabling element), poverty, and the environment.Moreover, transport is mainly an intermediategood, facilitating the production of final goodsand services that satisfy human needs. As suchit is an essential element of a city developmentstrategy, but not a freestanding one. It is only inthe integration of urban transport strategy withother sector strategies in responding to the prob-lems and opportunities of development thattransport strategy adds its value.

The critical point is that such integrated, strate-gic thinking should be realistic and should beconvertible into action plans and programs withinthe capability of the cities. Urban transport is asector more than normally subject to such dis-connect between vision and reality. New roadconstruction in the absence of a balanced urbandevelopment program that includes demandmanagement, public transport provision, andsupporting land-use policies may not improvetraffic or environmental conditions. Low-fare poli-cies for public transport, in the absence of a real-istic understanding of implied resource needsfor implementing this social strategy, may actu-ally cause deterioration in service. For that reasonthe review tries to balance a concern for the waytransport fits into the strategic vision of the citydevelopment strategy with a concern for betterways to deliver transport, a matter that is far fromtrivial.

The first part of this volume considers how urbantransport can be used as an instrument of urbandevelopment and poverty reduction. The strat-egy has two main thrusts. First, as discussed inchapter 2, “Urban Transport and City Develop-ment,” poverty may be reduced through the con-tribution that transport makes to the efficiencyof the urban economy and hence to the overallgrowth of incomes. Second, as discussed in chap-ter 3, “Urban Transport and Poverty Reduction,”urban transport policies can be focused morespecifically on meeting the needs of the poor.The inability to afford good transport service isnot the only transport-related aspect of the qual-

CONTEXT FOR THE REVIEW 3

ity of life of poor people, however. Chapter 4,“Transport and the Urban Environment,” con-siders the urban environment and shows that thepoor are particularly vulnerable to transport-related air pollution, while chapter 5, “UrbanTransport Safety and Security,” considers prob-lems of personal safety and security in transport.

The second part of this volume considers howthe objectives (to develop a better commonunderstanding of urban transport problems, withparticular reference to the poor, and to articulatea strategy to address them) can be pursued usinga range of instruments. Chapter 6, “The UrbanRoad System,” considers the provision and man-agement of road infrastructure. Chapter 7, “PublicRoad Passenger Transport,” discusses road-basedpublic transport, including the role of the informalsector. Chapter 8, “Mass Rapid Transit,” consid-ers the role and limitations of mass transit.Chapter 9 , “The Role o f Nonmotor i zed

Transport,” pays special attention to nonmotor-ized transport, which plays a very important, butoften neglected, part in meeting the needs ofthe poor (and increasingly the less poor).Chapters 10, “Urban Transport Pricing andFinance,” and 11, “Strengthening UrbanTransport Institutions,” address two commonareas of weakness identified in the precedingchapters: the issues of pricing and financing ofurban transport, and the institutional arrange-ments for the sector. Finally, in chapter 12,“Meeting the Development Challenges: HowCan the Bank Contribute?” we consider impli-cations for the instruments and lending strate-gies of the Bank.

NOTE

1. All dollar amounts are current U.S. dollars,unless identified otherwise.

URBAN TRANSPORT AND CITY EFFICIENCY

In this chapter we consider the impact of urbantransport on the development of the city econ-omy. In most developing countries the urbansector accounts for at least 50 percent of the grossnational product (GNP); in some countries thatnumber is over 70 percent. Cities in developingcountries often devote 15 to 25 percent, andsometimes much more, of their annual expendi-tures to their transport systems. Between 8 and16 percent of urban household income is typi-cally spent on transport, although this can alsorise to more than 25 percent for the pooresthouseholds in very large cities. About one-thirdof all city infrastructure investment need is for thetransport sector. Despite recent developmentsin private sector involvement in transport infra-structure finance, most of this investment willhave to come through the city budget.

Urban population is expanding at more than 6percent annually in most developing countries. Inmany formerly rural economies, such as China,because of the need to decrease the number ofpersons dependent on agriculture and to improveproductivity in rural areas, urbanization is viewedas a prerequisite of growth. Within a generationmore than one-half of the developing world’s pop-ulation will live in cities. This implies an increase

of 2 billion—equal to the present-day total urbanpopulation of developing countries.1 The numberof megacities—cities with over 10 million inhabi-tants—is expected to double, with three-quartersin developing countries. Some growth will be inhigh-density peri-urban settlements outside therange of the existing urban facilities and authori-ties. Much growth is likely to consist of urbansprawl, which militates against adequate publictransport service supply, encourages auto depend-ence, and hence reduces accessibility to employ-ment and to urban facilities for the poor and verypoor. It is therefore important to explore possibil-ities of improving the economic performance ofcities by better integrating transport with otheraspects of city development strategy (CDS).

Cities exist because of economies of agglomer-ation associated with industrial and trade activi-ties. The “advanced” sectors are located there,and labor productivity is typically higher in citiesthan in rural areas. The dominance of large anddense capital cities in many developing coun-tries suggests that these advantages continueup to megacity size.

Within these cities motorized road transport is themain mode of movement. While longer-distancemovements of goods and passengers may makesignificant use of other modes, and nonmotorizedtransport (NMT) may perform an important role

URBAN TRANSPORT AND CITY DEVELOPMENT 5

URBAN TRANSPORT AND CITY DEVELOPMENT2Deteriorating transport conditions associated with urban sprawl and increased

motorization are damaging the economy of large cities. Structural policies such as

well-planned transport infrastructure expansion, planned deconcentration,

comprehensive management of land-use structure, or liberalization of land

markets can help, but they require careful coordination of transport policies

within a broader city development strategy.

6 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

in short-distance movements of passengers (andin some cities, of freight), most large cities in theworld that are not dependent on mechanized roadtransport for the majority of internal freight andpassenger movement are poor, relatively unpro-ductive, and wish to change their situations. Of allcities, megacities have the highest travel times, thegreatest congestion, and the most polluted envi-ronments. The strategic quandary, particularly incountries where the capital city dominates, is howto retain the economic benefits of city scale whilelimiting the deterioration of transport performancethat may be associated with size and density.

Particularly in Asia, this deterioration of transportperformance appears to be generating a rapidgrowth of motorized two-wheelers that are fasterthan either bicycles (because of their power) orbuses (because of their personal nature, thus theirmaneuverability). The new two-wheelers are soinexpensive that even the relatively poor can affordthem. For example, a recent study of Delhi, India,showed that with an average per capita income ofless than $2,000 per year, over 80 percent of house-holds have motorized vehicles, mostly two-wheel-ers. They offer personal motorized mobility, albeitpresently at a high environmental and accidentcost (see chapter 5), although technology exists toclean them up very substantially at little extra cost.Moreover, there is also evidence that they makemore effective use of road space per person thaneither bicycles or private cars. In the short term,therefore, there does appear to be a rather differ-ent development path available for the developingcountries involving greater personalized mobilitythan was available in the industrialized countriesat equivalent income levels in their development.The long-term strategic question is whether motor-cycles are simply going to be viewed as a transi-tion step to an unsustainable level of privateautomobile ownership and use, or whether bygood traffic management and segregation theycan be sustained as the core of a more mobile, butsafe and sustainable, urban transport system.

In addressing that quandary, it must be recog-nized that cities differ greatly in economic, social,

and spatial characteristics. Moreover, any indi-vidual city will change its characteristics over time.We cannot hope to produce a simple blueprintfor the development of urban transport systemsthat is appropriate for all cities at all times.Nevertheless, although each city has its ownpeculiarities, four characteristics stand out asexplanations of transport differences:

a. Income. Vehicle ownership is primarilydependent on income, in developing as inindustrialized countries. Though rich coun-tries tend to have more road infrastructurethan do poor countries2—and at the nationallevel, paved roads tend to be undersuppliedin countries with low and middle per capitaincomes3—the growth of urban road spacewith income is likely to be slower than that oftraffic volume with income. Hence, unlessvehicle use is dramatically restricted, as ithas been in Singapore, traffic levels and con-gestion are likely to increase with income.

b. Size and size distribution. As city size, andparticularly spatial extent, increases, so typ-ically do the average length of commute,the level of traffic congestion, and the envi-ronmental impact of road traffic. Megacitieshave some of the worst problems of urbanpoverty, as well as the worst problems ofurban transport.4 This is accentuated in coun-tries that are dominated by their capitalcities.5

c. Political history. The form of modern citiesinevitably reflects their historical transitionbetween economic and social systems. Themost notable are those differences betweenformer socialist planned cities, many of whichhad widely dispersed pockets of high-den-sity residences served by mass transit, andthose cities where market forces played agreater role in shaping land use.6 In particu-lar, the transition economies combine rap-idly increasing motorization with a rapidlydeclining fiscal capability to support their tra-ditionally extensive public transport systems.

URBAN TRANSPORT AND CITY DEVELOPMENT 7

d. Population growth rates. Rapidly growingcities are distinct for two reasons: theyappear to have above-average car owner-ship rates in relation to income for thenational average income levels, and theytend to have below-average proportions ofland space devoted to circulation. Togetherthese militate for high congestion.

These influences clearly overlap and interact.Abstracting from the issue of city size, they giveus a taxonomy of city types into which majorcities can be divided (table 2.1), but which tosome extent explains the type of public trans-port systems that they have acquired. For exam-ple, high-income countries are highly motorizedand congested, but also tend to be more ableto afford rail-based mass transit systems. Wheregrowth has been very rapid, the developmentof mass transit is less likely to have kept pace.Where population growth has been slower, andparticularly for the cities in formerly centrallyplanned economies that have suffered stag-nating incomes, the probability of there beingmass transit systems is greater than incomealone would suggest. These differences betweentypes of cities, and the influences which causethe differences, should be borne in mind wheninterpreting the more generic discussions thatfollow.

THE ECONOMIC IMPACT OF POORURBAN TRANSPORT The pressures on urban transport systems areincreasing in most developing countries as partof the process of growth. Motor vehicle owner-ship and use are growing even faster than pop-ulation, with vehicle ownership growth rates of15 to 20 percent per year common in some devel-oping countries. The average distance traveledper vehicle is also increasing in all but the largest,most-congested cities. This growth exceeds theability to increase road space, and the majorimpediment to the efficient working of the urbaneconomies in large-size cities, and particularly inmegacities, is the level of road traffic congestion.Travel speeds are decreasing and the travel envi-ronment for pedestrians and people-poweredvehicles is deteriorating. Downtown weekdaytraffic speeds are reported to average 10 kilo-meters per hour (km/h) or less in Bangkok(Thailand), Manila (Philippines), Mexico City(Mexico), and Shanghai (China); 15 km/h or lessin Kuala Lumpur (Malaysia) and São Paulo (Brazil).It is estimated that congestion increases publictransport operating costs by 10 percent in Rio deJaneiro (Brazil) and 16 percent in São Paulo. Ofthe 16 developing-country cities with populationsof more than 4 million, 5 of them (Bucharest,Romania; Jakarta, Indonesia; Kinshasa, Republicof Congo; Lagos, Nigeria; and Manila) cited aver-

TABLE 2.1 A CATEGORIZATION OF CITY CIRCUMSTANCES

Source: Authors.

Po

pul

atio

n g

row

th

Hig

hLo

w

Income/motorization rates

Low High

Singapore Formerly centrallyplanned

Dhaka, Bangladesh Bangkok; Manila; MarketHong Kong, China

Samarkand, Uzbekistan; Moscow, Russian Federation; Formerly centrally Almaty, Kazakhstan; Warsaw, Poland;

plannedBishkek, Kyrgyz Republic Budapest, Hungary

Dakar, Senegal; Prague, Czech Republic; MarketNairobi, Kenya Buenos Aires, Argentina

8 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

age one-way commute times of one and one-quarter hours or more (UNCHS 1998). Growth ofmeasured gross domestic product (GDP) is alsoreduced by freight congestion, delays and unpre-dictability, difficulties of conducting business, andincreasing signs of disarticulation of the labormarket in some large cities such as São Paulo,Mexico City, and Manila. All this is occurringdespite the fact that motorization is still at a rel-atively early stage in most developing and tran-sitional economies; most developing countrieshave fewer than 100 cars per 1,000 people, com-pared with 400 or more per 1,000 people in thericher industrialized countries.

Furthermore, most transport-originated air pollu-tion, as well as nonbusiness time lost to conges-tion, is efficiency reducing but is not directlyreflected in GDP statistics. The safety and secu-rity of travelers is also diminishing in many largecities. Some of these impacts can be, and havebeen, valued in monetary terms. Table 2.2 pres-ents a summary of some estimates of external costsof road transport at national and regional levels.Recent World Bank estimates suggest that thetotal economic damage of air pollution representsup to 10 percent of GDP in polluted cities such asBangkok, Kuala Lumpur, and Jakarta.7 For sixdeveloping-country cities with a total populationof over 50 million (Mumbai, India; Shanghai;Manila; Bangkok; Kraków, Poland; and Santiago,Chile), World Bank estimates show the costs ofparticulates and other vehicle emissions (exclud-ing lead) as equivalent to 60 percent of the importcost of gasoline and over 200 percent of the importcost of diesel.8

SOURCES OF DECLINING URBANTRANSPORT PERFORMANCE

It is sometimes presumed that the deterioratingstate of urban transport in many developing coun-tries has been caused by relatively higher levelsof motorization with respect to income levels thanare experienced in the industrialized countries.The evidence does not support that proposition.

In terms of the relationship between income andcar ownership, the developing countries are fol-lowing a pattern very similar to that followed bythe industrialized countries, as figure 2.1 shows.Most developing countries fall in the develop-ment track shown for France, Japan, Spain, andthe United Kingdom. Only Argentina, Brazil,Mexico, and some of the transition countries ofEastern Europe have higher car-ownership-to-income ratios than the industrialized countriesexperienced. Chile, the Republic of Korea, thePhilippines, and Thailand all have lower nationalrates, but all have highly congested capital citiesthat have much higher incomes and (contrary toexperience in most Western industrialized coun-tries) much higher car ownership rates than thenational average.

The problems of the developing countries thusdo not generally seem to result from motoriza-tion occurring at lower per capita income levelsor at higher rates of income growth than thatexperienced in the earlier growth of the indus-trialized countries. Nevertheless, there are somerespects in which the present situation doesappear to differ from that of the industrializednations at a similar stage in their income growth:

• High concentration of national population,economic activity, and motorization itself inone or a very few major cities that areexpanding rapidly in size and population

• Inadequate quantity and structure of roadinfrastructure, often associated with rapidpopulation growth

• Poorly developed institutional, fiscal, and reg-ulatory arrangements at the municipal level.

LONG-TERM DYNAMICS OF URBANECONOMIC STRUCTUREThere is also a long-term dynamic interactionbetween transport and the nature of the cityeconomy. Cities have economic cores explainedby various forms of agglomeration economies,which are often based on a traditional industrialor trading base. Those employed in the citycenter choose their places of residence by trad-

URBAN TRANSPORT AND CITY DEVELOPMENT 9

TAB

LE 2

.2

ES

TIM

AT

ES

OF

EX

TE

RN

AL

CO

ST

S O

F R

OA

D T

RA

NS

PO

RT

(AS

PE

RC

EN

TAG

E O

F N

AT

ION

AL

AN

D R

EG

ION

AL

GD

P)

Po

lluti

on

Land

Acc

iden

ts,

Rev

enue

N

et

Co

untr

y o

r R

oad

an

d

Co

n-ne

t o

f Su

bfr

om

ro

ad

sub

-ci

tyYe

arSo

urce

cost

sp

arki

ngg

esti

on

insu

ranc

eN

ois

eLo

cal a

irG

HG

sO

ther

tota

lus

ers

tota

lO

ther

sTo

tal

Uni

ted

Sta

tes

119

89W

RI

1.64

a1.

56—

1.00

0.16

0.18

0.50

—5.

04a

5.04

0.46

5.5

Uni

ted

Sta

tes

2b19

90N

RD

C1.

25a

0.43

–1.7

40.

191.

710.

052.

09–3

.83

—0.

075.

69–8

.84

a5.

69–8

.84

0.78

–2.6

16.

47–1

1.45

Uni

ted

Sta

tes

319

91Le

e1.

762.

41—

0.24

0.19

0.73

—0.

265.

590.

884.

710.

875.

58

EU

1E

arly

19

90s

EC

MT

1.75

—0.

742.

400.

300.

600.

50—

6.30

1.67

4.63

—4.

63

EU

2E

arly

20

00s

EC

MT

1.49

—0.

751.

200.

300.

150.

47—

4.36

1.67

2.69

—2.

69

Uni

ted

K

ing

do

m19

93C

SER

GE

0.24

—3.

030.

46–1

.49

0.41

–0.4

93.

120.

02—

7.28

–8.3

92.

604.

68–5

.79

—4.

68–5

.79

Mex

ico

City

1993

Och

oa

—0.

082.

562.

32c

—0.

64—

—5.

60—

5.60

—5.

60

Pola

nd19

95IS

D1.

14—

0.30

1.60

0.10

0.30

——

3.44

2.81

0.63

—0.

63

São

Pao

lo19

00IR

BD

——

2.43

1.11

—1.

55–3

.18

——

——

5.09

–6.7

2—

5.09

–6.7

2

Bue

nos

Aire

s19

95FI

EL

0.73

—3.

420.

5–2.

00d

—0.

97—

—5.

62–7

.12

1.01

4.61

–6.1

1—

4.61

–6.1

1

Ban

gko

k19

95M

isc.

——

1.00

–6.0

02.

33—

2.56

——

5.89

–10.

89—

5.89

–10.

89—

5.89

–10.

89

Sant

iag

o19

94Ze

gra

s1.

371.

921.

380.

940.

152.

58—

—8.

351.

646.

71—

6.71

Dak

ar19

96Tr

acte

bel

——

3.37

0.16

–4.1

2—

5.12

——

8.65

–12.

61—

——

8.65

–12.

61

— N

ot

avai

lab

le.

No

te: G

HG

s =

gre

enho

use

gas

es; E

U =

Eur

op

ean

Uni

on.

In m

ost

cas

es, c

ong

estio

n co

sts

are

calc

ulat

ed in

co

mp

aris

on

with

eith

er a

free

-flo

w s

pee

d o

r an

“ac

cep

tab

le”

traf

fic p

erfo

rman

ce, a

ndno

t w

ith a

cal

cula

ted

“o

ptim

um”

leve

l of c

ong

estio

n. T

he c

alcu

late

d v

alue

s m

ay t

hus

ove

rest

imat

e w

hat

it w

oul

d b

e ec

ono

mic

to

elim

inat

e.a.

Ro

ad c

ost

s g

iven

net

of r

even

ues

fro

m r

oad

use

rs.

b.

Car

s o

nly.

c. G

ross

of i

nsur

ance

co

mp

ensa

tion.

d.

Cal

cula

ted

on

natio

nwid

e b

asis

and

gro

ss o

f ins

uran

ce c

om

pen

satio

n.So

urce

: Will

oug

hby

2000

a.

10 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

ing off increased travel costs against the lowerland costs (and greater availability of space andamenities) as the distance between places of res-idence and employment increases. As both city-center congestion and incomes increase, peopleare willing to pay more for space and amenities,so they travel farther to live where land is lessexpensive and home-based movement easier.Thus the area of the city increases. Paradoxically,to avoid road congestion, people move to loca-

tions in which they become increasingly depend-ent on the car. This trend is accentuated by invest-ments in radial trunk route capacity and bytechnology improvements that increase speedand reduce cost.

Similar considerations motivate firms. As the citygrows and its economic base diversifies, thosewhich need more space—often the larger export-oriented companies engaged in modern assem-

FIGURE 2.1 MOTORIZATION AND INCOMES: GROWTH OF CARS VERSUS GROWTH OF

PER CAPITA INCOMES IN FRANCE, JAPAN, SPAIN, AND THE UNITED KINGDOM, 1950S

TO 1995; AND THE RELATIVE POSITIONS OF SELECTED OTHER COUNTRIES, 1995

0

50

100

150

200

250

300

350

400

450

500

0 2 4 6 8 10 12 14 16 18 20 22 24 26

CZECH REPUBLIC

POLAND

ARGENTINA

MEXICOROMANIA

UKRAINE

MOROCCO

PHILIPPINES THAILAND

CHILE

RUSSIABRAZIL

MAURITIUS

KOREA

ISRAEL

SINGAPORE

HONG KONG,CHINA

GNP PER CAPITA (1995, THOUSANDS OF U.S. DOLLARS, PPP BASIS)

PA

SSE

NG

ER

CA

RS

PE

R T

HO

USA

ND

PO

PU

LATI

ON

OTHER COUNTRIES (1995)

SPAIN (1954–95)UNITED KINGDOM (1951–95)

JAPAN (1956–95)FRANCE (1951–95)

Source: Authors.

URBAN TRANSPORT AND CITY DEVELOPMENT 11

bly-line production technology—may flee fromthe old, high-rent, congested city center to loca-tions with less expensive land cost and betterexternal access to ports or intercity trunk roadsystems, while firms in trades and services maystay in the center to maintain access to cus-tomers.9 In many industrialized countries, this hasalready led to a relocation of freight move-ment–intensive activities to peripheral locations.The existence of a limited number of major radialtransport routes may lead to star-shaped pat-terns. There may be several subcenters along anyradial; where radials intersect with external rings,new nodes of high commercial density emerge.Despite this the old city center survives, retain-ing the highest-value commercial activities.

In richer industrialized countries, urban sprawl ischaracterized by unbounded outward spread ofdevelopment from the urban core, at low den-sity, often “leapfrogging” areas of undevelopedland to enter new jurisdictions competing fordevelopment.10 When employment follows res-idential development, it creates the phenome-non of the “edge city.”11 It is generally agreedthat sprawl increases the public and private costsof infrastructure per residence, while draining thefiscal capability of the traditional core causingdecay of the core’s infrastructure and decline ofits services. In transport terms it is generallyagreed that this increases trip lengths (even whenemployment is also decentralized) and autodependence, although it does not necessarilyincrease household travel expenditure, trip times,or overall congestion. Sprawl fosters a spatialmismatch between the places of work and resi-dence of the poor. Despite these disadvantages,suburbanization appears inexorable in higher-income cities.12

Not all of these features of sprawl are replicatedin the municipal development process of transi-tional and developing economies. In many of theformer socialist cities, leapfrogging was part ofthe planned structure, with high-density enclavesin peripheral areas surrounded by undevelopedland and linked to centralized employment by

high-capacity public transport links. The newsprawl in these economies tends to take resi-dences even farther out, however, and away fromthe traditional high-capacity public transport axes.In many of the rapidly growing developing cities,it is the poor rural-to-urban migrant who is locatedin the peri-urban settlements; land-use policiesmay even counter inner-city densification. Theperi-urban settlements are too low income to beserved by private automobiles, so that their devel-opment is associated with increases in householdtravel time and household budgets.

From the individual or company point of view,this outward shift of activity is an economicallyrational adjustment to increase accessibility, eitherto the amenity of space in the case of householdsor to markets and suppliers in the case of firms.In making decisions to change locations, how-ever, neither individuals nor firms need to takeinto account the indirect effects that their deci-sions are having on others. If transport prices areless than real costs (because they do not takeinto account congestion or environmental effects)or if infrastructure is provided below cost in newperipheral locations (because connection chargesand impact fees are too low), then the city willsprawl farther and faster than is economicallyoptimal. Some administrative action will then benecessary to curb this distortion. Planning inter-vention may also be necessary at the micro orlocal level, where new industrial developmentsmay have adverse effects on the amenity of exist-ing residential activities for which they are notcharged, with the consequence that the mix ofactivities (or environmental protection in mixedactivities) is suboptimal.

URBAN TRANSPORT IN CITYDEVELOPMENT STRATEGIES

The process of strategy development will varyfrom city to city. It is likely to require collectivevision for the city shared by city government andthe major stakeholders in civic society; an agreed-on strategic framework for realizing that vision;

12 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

a technical capacity to convert the strategy intopractical actions; and a fiscal and financial systemable to mobilize and allocate the necessaryresources efficiently (box 2.1).

Our starting point is the common observation thattransport is not usually demanded for its own sake,but that the demand for transport is derived fromthe demand for final consumption goods and serv-ices and for the raw materials and intermediateproducts that go into the production of final goodsand services. The implication is that transport prob-lems have two quite different generic types of solu-tion. The first type, discussed below, involvesinstruments internal to the transport sector to makethe sector more efficient. This may involve largeinfrastructure investments but may also involveimprovements in the management of infrastruc-ture to make it more productive.

The second type of solution is to operate on thesectors that generate the demand for transport.Most attention is usually given to the location of

activities, discussed below. There is a range ofother structural impacts that result from the roleof transport in servicing developments in othersocial sectors, such as health, education, and soon. Hence urban transport policy needs to beintegrated with other sector developments, notonly at the activity planning level but also in munic-ipal budgetary arrangements, as discussed below.

DEVELOPMENT STRATEGIES FORMAKING URBAN TRANSPORTMORE EFFICIENT

When preparing development strategies, anumber of issues must be examined, includingexpansion of the road infrastructure, urban freightpolicies, and the roles of the private sector.

ROAD INFRASTRUCTURE EXPANSION High levels of traffic congestion certainly give animpression of inadequacy of urban road infra-structure in many developing countries. It is well

BOX 2.1 CITY DEVELOPMENT STRATEGIES

A CDS is an action plan for equitable growth in cities, developed and sustained through partici-pation, to improve the quality of life for all citizens.

The goals of a CDS include a collective city vision and action plan aimed at improving urban gov-ernance and management, increasing investment to expand employment and services, and sys-tematic and sustained reductions in urban poverty.

It is not intended to substitute for integrated master plans, general land-use plans, or even invest-ment plans. Rather, it forms the basis for planning of land use, transport, and other sectoral needs,and for setting policy, resource allocation, and investment priorities.

Cities such as Bilbao (Spain), Rio de Janeiro, Sydney (Australia), Bangalore (India), and Yokohama(Japan) have successfully guided their growth with CDSs through a wide variety of approaches,with the lead being taken by the city itself, and with the urban poor and local business leadersactively engaged within a wider participatory process. In this way each of the cities secured abetter alignment and more efficient mix of public and private resource commitments. The Bankand its partners, through the Cities Alliance partnership, is engaged in more than 50 cities that arecurrently preparing CDSs.

Source: Tim Campbell and Peter Palesch, World Bank.

URBAN TRANSPORT AND CITY DEVELOPMENT 13

known that the 10 to 12 percent of land spacedevoted to all forms of road rights-of-way in themajor cities in Asia13 falls far short of the 20 to 30percent common in U.S. cities. It is clearly neces-sary to provide an adequate basic road networkand to extend it as the city expands in space.Primary rights-of-way or easements for these roadsshould be acquired at the outset of any devel-opment for future extensions of main arteries.This will give clarity for all parties about the broadfuture shape of the city and allow the land marketto operate more effectively to support rationalland-use patterns. Moreover, the character andcapacity of this infrastructure provision needs tobe tailored to the nature and density of theplanned or anticipated developments. It is par-ticularly important for secondary cities to use trans-port infrastructure to structure urban growth.14

Nevertheless, simple statistical comparisonsshould be viewed very cautiously. In fact, onlyabout 13 percent of land space is devoted toroads in London (out of about 18 percent devotedto total transport); the figures for Paris andMoscow are not much higher. There are severalreasons for these relatively low figures. Much railmovement space is underground in the Europeancities. Land-occupation density also has an impor-tant influence on the performance of the transportinfrastructure network. The structuring of the roadnetwork is also very important. It is necessary toprovide for local distribution of traffic as well asfor longer-distance trunk movements within andbetween towns. These functions do not mix well,and a given amount of road space will always givebetter performance if it is organized hierarchi-cally to try to separate functions. Some cities,such as Bangkok and Manila, suffer particularlybadly from the absence of an appropriate struc-ture of local distribution capacity. Thus it is themanagement and use of the space devoted totransport, rather than the simple proportion ofland devoted to roads, that is critical to systemperformance.

Furthermore, even if the proportion of spacedevoted to movement in an already highly con-

gested city or in a megacity is low, that does notmean that it can escape its problems simply bybuilding more roads. First, once the city fabric isestablished, it becomes increasingly expensiveand both socially and environmentally disturb-ing to superimpose substantial additional roadinfrastructure. Second, where congestion isalready suppressing demand, increasing capac-ity may simply generate such a large amount ofextra traffic that the congestion-reduction effectsare much lower than anticipated.

These considerations have some technical impli-cations for the appraisal of investments in extraroad infrastructure. Taking the economic and envi-ronmental impact of new traffic generation intoaccount will reduce the benefits attributed to thereduction of congestion for existing traffic, offsetto some extent by the marginal benefit of thenewly generated trips. Similarly, the alternativepossibilities—improving the efficiency of exist-ing infrastructure by traffic management, restrain-ing traffic by demand management, and shiftingtraffic from private to public transport—shouldalways be considered as the basis for assessingthe need for additional capacity.

URBAN FREIGHT TRANSPORT POLICIESFreight transport attracts relatively little atten-tion in the methodology of conventional Westernurban transport planning, mainly because it doesnot contribute much to the peak-hour flows,which are the predominant generators of con-gestion and which drive both physical investmentand traffic management responses to conges-tion. However, it does tend to attract specificattention where freight vehicles impinge on theamenities of residential areas. In the short termthis is often addressed by restraining the move-ment of freight vehicles. In the longer term theseproblems tend to be addressed through zoningand land-use controls. Consequently, both lightindustry and warehousing tends to move to theperipheries of the cities. Given the increasedimportance of reliable delivery time in integratedlogistic systems, and the more footloose natureof much modern light industry that makes it easier

14 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

to move to alternative locations, this focusedstrategic treatment of freight movement tendsto channel freight movements in ways that aremutually acceptable to those concerned withlocal congestion and the environment, and thoseconcerned with efficient freight movement. Evenin the great port cities of the world, the shift ofport facilities downstream has usually shifted con-centrated freight traffic flows out of the citiesrather than into them.

Very different considerations apply in manydeveloping-country cities.15 Ports, industry, andcommodity markets often remain in their his-toric city-center locations. Roads in these areasare often narrow and encumbered, requiring theuse of more, and smaller, vehicles for freightflows, including very small nonmotorized vehi-cles—such as handcarts, bicycles, and rick-shaws—in many places. Peripheral infrastructure,which attracts much freight-intensive activity outof the center of developed-country cities, is lesswell developed. Freight vehicles thereforeaccount for a larger proportion of urban roadtraffic flows, freight transport performance tendsto be inferior, and the perceived congestion andenvironmental impact of freight transportgreater.

Many of the most successful developing coun-tries have achieved rapid growth through export-led industrial development. This development isoften based, initially, on the exploitation of inex-pensive domestic labor. As incomes rise, how-ever, it depends more on the labor force’s qualityand its ability to participate efficiently and flexi-bly in global manufacturing and distribution sys-tems. Reliable transport, both within cities andin connections to international networks, is crit-ical to this development.

For the most part, logistic networks are entirelyprivate sector, and often led by multinationalcompanies. However, there are some importantexternal requirements for these networks’ devel-opment. An open domestic transport market,with freedom for modal integration and good

seaport and airport facilities, is central to theserequirements. This transport market must besupported by good telecommunication facili-ties. These elements are to be found in the rap-idly industrializing economies, and are notablyabsent in the stagnating ones. City growth thusdepends not only on actions that can be taken atthe city level but also on the support of a nationalgovernment committed to liberalizing freightmovement.

That still leaves some important supporting policyrequirements for economic growth at the munic-ipal level. Development planning, supported byland-use control, can foster well-located indus-trial development both in relation to externaltransport links and in relation to domestic labormarkets, thus avoiding both local traffic conges-tion and divisive environmental impacts. A goodplanned example of this approach is Curitiba,Brazil. The effective protection of the city envi-ronment from the impacts of industrial and port-related traffic through the combination of roadinfrastructure and industrial location policies isalso one of the most notable features of theDutch planning system.

Many large cities draw their wealth from com-mercial as much as from industrial activities. Hereagain the combination of a liberal commercialenvironment with good internal transportationand telecommunication links is critical to theattraction of the city to international business.

THE ROLE OF THE PRIVATE SECTOR Responsibility for urban transport is being decen-tralized to the cities in many countries. In thisdecentralization process, the expenditure respon-sibilities of municipalities tend to expand muchmore than the intergovernmental transfer of finan-cial resources. Only a small fraction of the nec-essary resources can come from the multilateraland bilateral agencies.16 Cities will therefore haveto broaden their financial bases in order to beable to finance the infrastructure that they want,as well as to support any social objectives theyare pursuing in controlling transport operations.

URBAN TRANSPORT AND CITY DEVELOPMENT 15

As a result cities increasingly look to the privatesector to provide facilities and services.

As far as transport services are concerned, it is esti-mated that at least 80 percent of all urban busservices provided around the world are now pri-vately owned and operated, including those oper-ated privately for public sector clients. The onusfor financing both rolling stock and, to a lesserextent, the supporting infrastructure is thus trans-ferred to the private sector to be financed throughfares. There is increasing concern, however, aboutthe quality of service available in many unregu-lated regimes. As discussed in chapter 7, some ofthese concerns can be addressed effectively withcompetitively tendered franchising of services.That requires a clear legal and regulatory frame-work, as well as a strong public planning and publicprocurement capability. While the legal frameworkis typically provided through national legislation,the planning and procurement is essentially a localresponsibility that depends heavily for its successon how well these arrangements relate to the restof the city development strategy (CDS).

Many municipalities harbor parallel expectationsthat transport infrastructure capacity can also beprivately financed. In many cases they have seensuccessful privatization or concessioning of power,water, and telecommunication utilities, andbelieve that roads and mass transit systems canbe similarly financed. Although some existingurban expressways and urban railways have beensuccessfully concessioned in a number of coun-tries (see chapter 7), new infrastructure is moredifficult to finance privately. In the road sector,the scope for private financing through tolls islimited by the need to be able to limit access. Inany case, the proportion of the network that canbe financed in that way is relatively limited.Shadow toll systems have been developed toextend the scope of private participation, butexperience is so far limited, and in any case thecharges of such systems remain on budget.

As far as urban rail systems are concerned, pureprivate finance has not yet been successful in a

developing country (though the Bangkok TransitSystem [BTS] could change this conclusion). Thefinancial difficulties arise partly due to under-charging of the competing, congested roadspace. Significant operational difficulties havealso been experienced in cities such as KualaLumpur, Manila, and Bangkok because of theirfailure to integrate the private systems effectivelywithin a comprehensive urban transport anddevelopment strategy. These issues are discussedfurther in chapters 8 and 10.

It is not the intention here to discourage increas-ing private sector participation in urban transportinfrastructure finance, but there are some gen-eral observations that can be made on the require-ments for success in utilizing private financeeffectively. First, purely opportunistic finance is tobe avoided. Unless the private developments con-form to a general structure plan, they may imposeunforeseen, and sometimes very significant, con-tingent costs on the public budget. Second, care-ful attention is required to individual contractdesign to ensure that the objectives of the CDSand the financial requirements of the private fin-anciers are effectively reconciled. Third, cities mayneed to consider carefully their priorities for pri-vate finance in order to strengthen their credit-worthiness to be able to access domestic andinternational capital markets. These considera-tions all highlight the need for any attempt tosecure private financing of transport infrastruc-ture to be integrated within a CDS, both withrespect to physical and to financial planning.

STRATEGIES FOR STRUCTURALCHANGE: LAND-USE AND TRANSPORT PLANNINGIn most countries local transport performance isbetter, and costs lower, in smaller rather than inlarger cities. In a number of developing coun-tries, of which Thailand is the extreme example,the dominance of the capital city is such thatthese differences are very large. Structural strate-gies to improve transport may therefore focus onthe distribution of activities among cities—par-ticularly through planned deconcentration—or

16 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

the distribution within cities through land-useplanning and development controls.

DECONCENTRATIONLimiting the further expansion of the major citiesby consciously promoting the development oflower-order urban developments is a beguilingstrategy. France and the United Kingdom haveboth adopted policies of controlling the capital,encouraging the provinces, and developingperipheral new towns. China has similarly beenseeking effective policies to discourage all ruralmigrants from going to the booming coastal cities.The case for limiting further capital-city growthbecause of intolerable transport conditions hasbeen argued even for a city of only 2 million inhab-itants, such as Nairobi.17 Given the transport prob-lems of many large cities, it is tempting to seedecentralized urban growth as a solution.

There are two major problems with that strategyfor attenuating the transport problems, and otherproblems, of the megacities. First, neither themagnitude of agglomeration economies nor thesignificance of environmental externalities areunderstood clearly enough to be able to makeany universal judgment about how far or how fastto push deconcentration. Second, attempts tocontrol concentration, either by land-use anddevelopment constraints in the megacities or byinducements to locate outside the megacities,have had limited impact and dubious benefit.18

Many capital-city authorities prove to be luke-warm in their support for such policies and tendto resist even the dispersion of some central gov-ernment activities.

Given these doubts, it is likely to be better toconcentrate on using macroeconomic policiesto allow markets to work more effectively in locat-ing activity, rather than to engage in strongadministrative redirection of economic activity.Protectionism and autarky appear to reinforcethe dominance of capital cities. Open marketsare likely to create a more-level playing fieldbetween core and peripheral manufacturing andbetween urban and rural activities. The quality

of intercity transport and communications canalso contribute to that level playing field. Forexample, it has been estimated that a 1 percentincrease in the share of GDP spent on govern-ment transport and communications investmentsis associated with a 10 percent reduction in pri-macy; barriers to internal trade reinforce pri-macy.19 Similarly, the removal of subsidies to themegacity—including transport subsidies—mightreduce megacity size without adverse social con-sequences.

PLANNING AND MANAGEMENT OFURBAN LAND USETwo extreme approaches to improving the struc-ture of activities within cities may be character-ized.20 Urban planners tend to define a preferredurban form, albeit based on extensive consulta-tion and consideration of transport demand,incorporated in a structure plan. This providesthe required physical framework for market forces,private sector investment, and public sector pro-grams for urban change and growth. The plancan be indicative and passive, or it can be activelypursued through public sector transport andurban regeneration investment. Economists, incontrast, tend to concentrate on applying an effi-ciency criterion to each policy decision, withoutprejudging the structural outcome.

Neither approach is sufficient in itself. Full inter-nalization of externalities, precluding the needfor any planning intervention, has not beenachieved even in the most sophisticated ofmarket economies, such as the United States.Moreover, the longevity of major infrastructureis such that conventional financial decisionmak-ing discounts the effects over much of its life. Onthe other hand, planning undoubtedly worksbetter if supported by, rather than workingagainst, economic incentives. Thus it is advisableto look at the effectiveness of both administra-tive and market-based instruments in the searchfor a strategy on land use and transport.

The possibility of manipulating urban land usefor transport policy purposes has been recently

URBAN TRANSPORT AND CITY DEVELOPMENT 17

incorporated into an environmental argumentfor densification, based on the undisputed factthat gasoline consumption per capita in cities ishighly correlated with the overall density of thecity.21 Proponents of this view argue that reduc-ing density increases trip lengths, makes publictransport less viable, encourages greater use of,and dependence on, the private automobile (autodependence), and hence generates more envi-ronmental impact per capita. They also point tothe adverse effects of increasing auto depend-ence on those without access to a private car,who find themselves progressively excluded fromaccess to economic and social activities.

In their simplest form, policies designed to pro-mote land-use concentration so as to save energyhave been challenged on the grounds that energyconsumption is but one input, and that it is inap-propriate to concentrate only on minimizingenergy consumption irrespective of its effects on

other aspects of the quality of life (particularlythe amount of residential space per person).22

This has also proved to be a very difficult area forpolicy implementation, even in sophisticated gov-ernance systems. There is much evidence to showthat where individuals have been free to choosethe location of their residence, suburbanizationhas continued as incomes increase, irrespectiveof land-use planning policies. This has shifted thefocus of research in the direction of obtainingbetter understanding of the determinants of thisbehavior and transferred the policy emphasis toother instruments, such as land-use mix, publictransport structure and quality, which operate ontravel choices in ways that more obviously respectindividual preferences.

Effective administrative control of land userequires, above all, the willingness and foresightto plan. One of the defining features of cities thatare widely believed to have been most success-

BOX 2.2 MATCHING LOCATIONAL CHARACTERISTICS AND USER DEMANDS

In the Netherlands a system called the “ABC” system is used to match the characteristics of landuses to characteristics of the transport network. Types of location and types of activity are both clas-sified at a central government level according to their traffic-generating characteristics and accord-ing to their need for accessibility by public transport or by private road transport. The distinctionbetween the locations should then be reflected in public transport planning and in parking restric-tions on “A” and “B” locations implemented at the provincial or municipal level.

The approach requires a hierarchical control to avoid competition between adjacent jurisdictionsfrom undermining the functional allocations, as well as parallel financing actions on public trans-port development to prevent new residential developments from becoming auto dependent. Ithas had a significant effect in areas directly controlled by the national government. From dis-persed sites throughout the country, the Ministry of Housing, Physical Planning, and the Environmentrelocated to a new office building directly adjacent to the central railway station in The Hague. Thisresulted in a dramatic drop in commuting by auto from 41 percent to only 4 percent, while com-muting by rail increased from 25 percent to 57 percent. Bus and tram use went from 9 percent to20 percent. Lower-level authorities appear to have implemented the system with varying degreesof commitment. Nevertheless, it has undoubtedly contributed to sustaining the public transportshare of trips to major administrative, office, and educational locations, while ensuring good roadtransport access for industrial and commercial plants.

Source: NEA Transport Research and Training 2000.

18 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

ful in managing the relationship between trans-port and land use (such as Curitiba, Brazil; Zurich,Switzerland; and Singapore) is the early existenceof an integrated land-use and transport structureplan in support of which a wide range of sectoralpolicies were employed. In many countries thisis hampered by the lack of appropriate institu-tions and consistent political leadership at themetropolitan level. The hitherto successfulattempt to transfer experience on strategic plan-ning and urban development from Zurich toKunming (China) through a twinning arrangementenjoys the active support of both Swiss andChinese central governments.23 Structure plan-ning also needs to be supported at the imple-mentation level by guidelines relating the natureof the activities being developed in any locationwith the transport facilities available to it. In theNetherlands national guidelines have been estab-lished to foster such consistency in land-use andtransport planning (box 2.2).

A second essential requirement is the existenceof the necessary technical skills to develop a planthat is comprehensive and internally consistent.In the absence of an internally consistent and fullyintegrated spatial policy, infrastructure invest-

ments and regulations often work at cross-pur-poses. For example, in some Indian cities atten-tion is given to the avoidance of congestion inexisting built-up areas and to the need to pro-vide more city-center parking space, while at thesame time developing “suburban” greenfieldsites (box 2.3). A more consistent policy mightbe to let parking space provision be entirely aprivate, market-driven, activity. This would changerelative costs of public and private transport, andmight actually improve the split between publicand private transport.

A third requirement for effective planning of theland-use and transport interaction is the ability toimplement land-use, public transport, and devel-opment controls in a coherent manner that con-sistently supports the planning objectives. Forexample, many Indian cities adopt an unrealisti-cally low permitted floor area ratio in central areas,which not only restricts the ability of developersto build office and retail buildings where there ismost demand (in the central business district[CBD] and around transport nodes), but alsopushes new residential construction to the periph-ery. Similarly, imposition of inappropriate setbackand coverage requirements tends to keep the

BOX 2.3 INTERNAL CONSISTENCY IN PLANNING

In Bangalore, India, land-use regulations prevent the densification and the creation of new officespace in the city center. However, the city is planning a major investment in a light rail transitsystem whose economic feasibility depends on the creation of new jobs in the current centralbusiness district (CBD), and therefore in the building of new office space. Simultaneously, newtechnological parks are built in distant suburban areas (out of range of the light rail lines) to try tocreate enclaves of adequate infrastructure and services, while the city-center infrastructure is leftto deteriorate. The Bangalore Revised Comprehensive Development Plan for 2011 plans for threeconcentric ring roads (beltways), which would further contribute to the dispersion of jobs. Themain issue is the inconsistency between different projects. This implies that the appraisal of eitheris dependent on the strategic priority that is given to the other.

Similar problems exist in Ahmedabad, where the imposition of low floor space ratios in the CBDprevents the renovation of buildings in the prime location, and hence contributes to the flight ofactivities to the suburbs.

Source: Bertaud 1999.

URBAN TRANSPORT AND CITY DEVELOPMENT 19

wrong space underoccupied compared with apolicy of having a strong position on public openspace. Even if effectively enforced, as in Seoul,Republic of Korea, greenbelt policies tend to exertstrong upward pressure on the price of housing.24

THE ROLE OF MARKETS IN LAND-USEDISTRIBUTION Even in an administrative planning context,market tendencies and values need to be rec-ognized in implementation. Redevelopment canonly happen when regulations allow it, when realestate transaction costs are low, and when themunicipality is willing and able to redesign andfinance improved infrastructure to accommodatenew types of land use and land densities.

Any reduction in unit transport costs will tend tohave the effect of reducing density, expandingthe spatial scale of the city, separating differentland uses, and, possibly, increasing total trans-port expenditure and transport resource use(including fuel). Any undercharging of transportcosts—whether it is for road use, parking, orpublic transport—will similarly accentuate thesprawl. Urban boundaries and greenbelts mayconstrain that outward pressure for a time, butthey tend to be jumped over without necessarilyleading to the establishment of new self-con-tained cities with a balance of employment andresidence. An “efficient pricing” approach tourban structure attempts to reflect costs—both oftransport and of land occupation—in prices, andto allow individual preferences on land use toreconcile the variety of tastes with regard to spaceand other forms of consumption.

There are important limitations to this approach.Given that the major distortions arise from under-pricing of environmental and congestion impacts,a quantification of those effects is required, at ahighly disaggregate level, accompanied by acharging mechanism to implement the priceregime. The same applies to properly internaliz-ing the “costs of sprawl” in land developmentcharges. Finally, it is necessary to handle the redis-tribution implications of the charging mechanism.

Even if these difficulties exclude the possibilityof sole reliance on market forces, it is still pru-dent to attempt improvement of land-marketoperation.25 In developing countries there is along agenda of actions necessary to facilitatedemand-driven land-use change, including:

• Clarification and recording of property rightsto generate security of tenure, provide abase for investments and borrowing, andenable efficient transfers of propertybetween owners

• Establishment of procedures for speedyadjudication of land invasions and informalacquisitions, and for assignment of propertyrights, especially in cities where much of thehousing is provided and developed by theinformal sector

• Replacement of existing types of title, whichoften provide only restricted rights, with afull ownership title that is recorded in asingle, open, registry containing full infor-mation about liens, mortgages, easements,and so on

• Incorporation of the full costs of ancillaryinfrastructure development in connectioncharges for on-site public utilities and feesfor trunk infrastructure

• Conversion of existing land-use controls intoa transparent zoning and building controlsystem, sensitive to demand signals

• Elimination of obstacles to recycling of pub-licly owned lands, particularly in formerlyCommunist countries, in which enterprisemanagers and local government officialsoften combine to prevent reassignment ofindustrial land in prime locations

• Broadening and deepening of financial mar-kets’ support for housing, with emphasis puton groups who would otherwise be unableto obtain appropriate accommodation.

Regulations should also be revised to avoid dis-tortion of land use. Legally required standardsfor building and site development (including min-imum plot sizes, setbacks, and parking provisions)should reflect the availability and affordability of

20 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

land. Regulations should be firmly enforced.Market distortions resulting from excessive inter-jurisdictional competition should be minimizedthrough national standards in order to limit localgovernment ability to grant special favors (suchas tax remissions or exceptional relaxation of reg-ulations) that would attract investors away fromother jurisdictions.

Public sector pricing and taxing practices are oftenthemselves the source of distortions. Sales andleases of lands owned or developed by publicauthorities should always be at full market value,and public utilities should set connection chargesreflecting actual costs rather than systemwide

averages, with any exemptions specifically tar-geted and funded. Infrastructure cost should beincluded in land price through transparent impactfees or in-kind obligations. Developers should berequired to cover the costs of neighborhood infra-structure and whatever expansions of public facil-ities are needed to maintain service levels forservices such as fire and police stations, storm-water infrastructure, schools, roads, and busstops.26 Usually this involves a substantial publicsector planning effort. However, where govern-ment provides an enabling framework, the pri-vate sector can be stimulated to undertake a fullycoordinated development of the land-use andtransport infrastructure (box 2.4).

BOX 2.4 THE MARKET AND LAND USE AND TRANSPORT INTEGRATION: THE TAMA

GARDEN CITY DEVELOPMENT

The Tama Garden City Development Project is widely viewed as a model of integrated land-useand transport development. The project, promoted by the Tokyu Railway Company, was plannedto transform a vast, hilly, and sparsely inhabited area into a community of some 5,000 hectareswith 0.5 million residents, as well as to construct a railway of 22 kilometers (the Den-en Toshi Line)passing through the newly developed area and linking it to central Tokyo. The first phase of therailway construction (14.2 kilometers) was completed in 1966, followed by a second phase (5.9kilometers), which started in 1967 and was completed in 1984. The total construction cost was 22billion yen ($200 million), 50 percent of which was financed by commercial loans and the rest fromthe Japan Development Bank. No direct government subsidy was provided. On completion of therailway, bus routes, largely operated by Tokyu, were rearranged to provide feeder services for railusers. Between 1959 and 1989, nearly 3,000 hectares were developed for a population of 440,000,and the Den-en Toshi Line carried about 729,000 riders per day in 1994.

Among the key features of this project was the use of “land readjustment” to assemble the landneeded to accommodate the railway and to develop real estate. Rather than acquiring all of theland, Tokyu organized landowners to form a cooperative that consolidates properties, redevelopsthem without transferring ownership, and returns smaller but fully serviced parcels to landowners.A unique aspect of this project was that Tokyu undertook the whole construction without charg-ing the cooperatives for the redevelopment works; in return, Tokyu acquired the reserved hous-ing sites after completion of the redevelopment. The success of the first readjustment phaseaccelerated the formation of cooperatives, thereby leading to large-scale area developmentwithin a relatively short time. Tokyu and its affiliated companies actively promoted the area’s devel-opment in a variety of ways in order to increase population and rail ridership, including sellingland, constructing housing, developing and attracting shopping centers, and inviting schools tolocate themselves within the development.

Source: PADECO 2000.

URBAN TRANSPORT AND CITY DEVELOPMENT 21

While a free land market can thus improve theefficiency with which land use is allocated andadjusted to changing economic conditions, thereis a potential tension between the operations ofthe land market and the desire to focus transportpolicies to benefit the poor. Transport invest-ments change the structure of land values. If thereis strong competition for the use of land andhighly concentrated ownership of land, rentsincrease and the benefits of transport improve-ments accrue to rich landowners rather than topoor land occupants. Focused transport invest-ments may, through this process, simply drivepoorer people out to other, less-expensive, loca-tions. This potential of perverse redistribution,further discussed in the next chapter, emphasizesthe need for transport to be part of a compre-hensive urban development strategy in order toprevent the benefits of transport improvementsbeing appropriated exclusively by the better-offinhabitants. Some mechanism to capture, for thepublic good, the land value increases resultingfrom public infrastructure investments is critical.

COORDINATING SECTOR POLICIESIN CITY DEVELOPMENTSTRATEGIES

The implications of transport’s being a deriveddemand are not confined to the level of generalspatial structure. There are also important impli-cations for the formulation of a range of socialsector policies and for institutional and financialplanning arrangements.

SOCIAL SECTOR POLICIESIn the provision of health services, the public sectordominates in most countries. Whether health serv-ice provision is formally a national or local respon-sibility, the local health authority typically has a highdegree of independence from the municipalauthorities. This can have a number of adverseeffects. First, decisions on the size and location ofhospitals and clinics tend to be taken in the inter-est of minimizing health sector costs or improvinghealth sector service quality. Considerations of

accessibility of facilities, both for patients and vis-itors, tend to be given lower weight. Second, evenin countries such as Bangladesh, where the burdenof road accidents on health facilities is very large,little weight seems to be given to designing med-ical strategies to improve on-site availability of med-ical services, or to combining sector interests togive road safety a higher priority.

In the education sector, there is an even widerrange of issues. Location of facilities is usuallyless of a problem, although in some cities, suchas Santiago, Chile, the combination of liberalschool choice policies with the location of themore attractive schools in higher-income areasputs very heavy demands on the public transportsystem. That problem is often accentuated bythe planned coincidence of the journey to schooland journey to work peaks, especially in higher-income countries where some parents take chil-dren to work by car. From a public transportoperational point of view, that problem tends tobe accentuated by the traditional (and some-times mandatory) provision of reduced fares forstudents. As long as the financial burden of thesefares falls on the transport budget, and there is noinstitutional channel through which the potentialtradeoffs can be examined and negotiated, theoutcome is likely to be suboptimal.

A similar set of problems relate to social secu-rity. In some countries there is a statutory or con-stitutional basis for free or reduced-fare travel forpensioners or the unemployed, without refer-ence either to any limitations on that right (off-peak only) or to the financing of it. Where thatapplies only to public sector transport providers,it tends to obscure the issues in deciding howbest to organize public transport services.Particularly in some of the republics of the formerSoviet Union, the proliferation of ostensiblysocially motivated concessions has played a sig-nificant part in the decline of public transportservice capability.

Public sector administrative activities can beused as a leading sector, particularly in capital

22 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

cities, where government employment is con-centrated. For example, moving activities fromthe former central district of Mumbai to new loca-tions farther up the peninsula may make a veryimportant contribution to improving the trans-port situation in the old city center.

INSTITUTIONAL AND FINANCIALPLANNING IMPLICATIONS Most of these problems of intersector strategycoordination have both an institutional and afinancial dimension. Institutionally they call forchannels of coordination, both between func-tions and between jurisdictions. It is not possi-ble to propose a blueprint institutional structureapplicable to all countries and all political sys-tems. The most critical point, however, is thatthere should be some effective integrating mech-anism or process through which the issuesbecome explicit and receive attention. Some rel-evant principles and models are discussed indetail in chapter 11.

In parallel with the need for an appropriate coor-dination of institutional responsibilities is theneed for coordinated financial planning. Chapter10 presents the argument for a flexible financialsystem for urban transport, allowing resourcesto be efficiently allocated between modes. Insome cases this may justify the establishment ofa multimodal urban transport fund. The rela-tionship between that arrangement, justified interms of intrasector efficiency of resource allo-cation, and the municipal budgetary allocationprocess needs careful structuring.

CONCLUSIONS: A STRATEGY FORURBAN TRANSPORT IN CITYDEVELOPMENT

It has been argued that cities are the engines ofeconomic growth in most developing countries,and that urban transport is the oil that preventsthe engine from seizing up. Unfortunately, dete-riorating transport conditions are already dam-aging the economy of many large cit ies,

particularly the megacities, worldwide. Becausedemand for transport is essentially a deriveddemand, urban transport must be viewed strate-gically as an integral component of the city econ-omy and hence of its development strategy.

The economic performance of the sector can beimproved by more careful attention to therequirements of freight transport and logistics,as well as by improvements to infrastructure,including privately financed infrastructure, whereappropriate. While expanding cities require ade-quate infrastructure, it is physically and eco-nomically impossible to escape from congestionby building roads in the densest cities. Broaderstructural approaches are also required.Deconcentration of activities can be encouraged,but is difficult to achieve. Planning and manage-ment of land-use structure is essential, but haspractical limitations. Liberalization of land mar-kets can help, but the direct impact is weakenedby the inability to internalize the external costsof development. Integration and coordination ofsector policies are also central to the more inte-grated development approach.

Although no single, simple, structural policy offersa complete solution to the transport-related prob-lems of urban growth, it is possible to identifysome elements of each policy that can be advo-cated as robust components of a strategy fortransport within an urban development strategy.These elements include:

• Elimination of policies favoring the capitalcity together with properly appraised invest-ment in intercity transport outside the cap-ital region

• Development of a structure-planning capa-bility as the basis for positive CDSs.

• Provision within structure plans of space fortransport infrastructure that would be ade-quate for immediate demand but also capa-ble of adaptation as the city grows

• Coordination of the planning and develop-ment of land use with that of transport infra-structure and services

URBAN TRANSPORT AND CITY DEVELOPMENT 23

• Encouragement of development-controlskills and practices at the city level

• Elimination of obvious price distortions inboth land and transport markets, includingthe introduction of congestion prices forroad use and full-cost connection chargesand impact fees for land development

• Improved road investment appraisal to takeaccount of the economic and environmentaleffects of induced traffic in assessing theneed for capacity expansion

• Strategic consideration of the benefits thatcould be achieved through traffic manage-ment, and demand restraint in the base-casefor road investment appraisal

• Coordination of transport sector policies withthe policies of the sectors that transport serves.

NOTES

1. World Urbanization Prospects (UnitedNations 1996), projects urban population of 1.9 in2000 and 3.7 billion in 2025. By that date it is pro-jected that there will be 500 cities of over 1 mil-lion inhabitants and 35 megacities of over 10million inhabitants.

2. Ingram and Liu 1999.3. Canning and Bennathan 2000.4. There is less agreement, however, concern-

ing the size of city at which these problems beginto dominate the advantages of agglomeration,or why some cities, particularly in developingcountries, continue to grow despite havingreached this situation.

5. City size distribution is very skewed in manyAsian and African developing countries, but notin most Latin American or Eastern Europeancountries. In Thailand the capital city is over 40times as large as the second city, and this ratio

(the “primacy index”) is above three for otherh ighly populated countr ies such as thePhilippines, Malaysia, and Indonesia (Karan1994).

6. Dutt and others 1994.7. Hughes and Lovei 1999.8. Lvovsky and others 2000.9. This pattern, well documented in industrial-

ized countries, has been shown to be equallyforceful in developing cities such as Bogotá, Cali(Colombia), and Seoul. See K.S. Lee 1989.10. Burchell and others 1998.11. Garreau 1991.12. Fouchier 1997.13. It is 11 percent, for example, in Bangkok andKolkata (India).14. Koster and de Langen 1998.15. For a comparison of the cities of Rotterdam(the Netherlands), Dhaka, and Nairobi, seeArcadis Bouw/Infra 2000.16. In the case of the World Bank, the require-ment of a sovereign guarantee is a particularimpediment for cities in countries whose gov-ernments are unwilling to give such guarantees.17. Howe and Bryceson 2000.18. For example, the Korean greenbelt policyfor Seoul appears to have produced perversedensity gradients, high housing costs and traveldistances, and arguably has militated particularlyagainst the interests of the poor.19. Krugman 1991.20. D. B. Lee 1999.21. Newman and Kenworthy 1989.22. World Bank 1996.23. Joos 2000.24. Dowall 1995.25. Elaborations of these points are to be foundin Dowall 1995.26. Pendall 1999.

URBAN POVERTY AND SOCIAL EXCLUSION

Poor households derive their standard of livingfrom a variety of activities, not all of which are mar-keted or assigned a monetary value. That stan-dard of living, and its security, depends not onlyon current income but also on the stock of assets,including the social and human capital, as well asthe money and physical assets, at the disposal ofthe household. Poverty is thus a multidimensionalconcept involving the lack of the social and cul-tural, as well as economic, means necessary toprocure a minimum level of nutrition, to partici-pate in the everyday life of society, and to ensureeconomic and social reproduction.1 In this gen-eral notion of poverty as “exclusion,” accessibil-ity is important, not only for its role in facilitatingregular and stable income-earning employmentbut also for its role as part of the social capitalthat maintains the social relations forming thesafety net of poor people in many societies.

Deteriorating urban transport conditions have aparticularly severe impact on poor people.2

Growing reliance on private vehicles has resultedin a substantial fall in the share of, and in somecases an absolute decline in the number of, tripsmade by urban public transport in many cities.Consequently there has been a decline in urban

public transport service levels. Sprawling land-consuming urban structures are making the jour-ney to work excessively long and costly, particularlyfor some of the very poor. Surveys of commutersin Mexico City have shown that 20 percent of work-ers spend more than three hours traveling to andfrom work each day, and that 10 percent spendmore than five hours.3 Poor people also suffer dis-proportionately from deterioration of the envi-ronment, safety, and security because they arelocationally and vocationally most exposed, andless able to afford preventative or remedial action.

TRANSPORT PATTERNS OF THEURBAN POOR

Poor people make fewer trips per capita than dothe nonpoor. The difference in total number oftrips per day per person is not usually extreme,falling in the range of 20 to 30 percent, thoughsome earlier studies have suggested much greaterdisparities.4 Consistent with the difference betweentrip rates of the poor and the nonpoor, averagetrip rates have also tended to increase over timeas income increases.5 In contrast, the compositionof the trip making of the poor and the nonpoordiffers very substantially. The nonpoor typicallymake two or three times as many motorized tripsper capita as do poor people, even when total trip

URBAN TRANSPORT AND POVERTY REDUCTION 25

URBAN TRANSPORT AND POVERTY REDUCTION3Poor people’s inability to access jobs and services is an important element of the

social exclusion that defines urban poverty. Urban transport policy can attenuate

this poverty, both by contributing to economic growth and by introducing a

conscious poverty reduction focus to infrastructure investment, to public

transport service planning, and to fare-subsidy and financing strategies. There is

a rich agenda of urban transport policies that are both pro-growth and pro-poor,

yet which are consistent with the fiscal capabilities of even the poorest countries.

26 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

rates are fairly similar. In most poor countries, pri-vate motorized vehicle trips are restricted to thewealthiest 20 percent of the population, with themotorcycle extending this down to those with aver-age incomes in middle-income countries.6 Asmight be expected, poor people’s journey pur-poses are more restricted, with journeys to work,education, and shopping dominating.

The burden of transport on household budgetsoften cannot be determined precisely. Incomesmay be difficult to establish, especially where thereis some payment in kind or where there are incen-tives not to disclose the total income. Householdexpenditure is therefore probably a better basethan is income, although it is believed that house-hold consumption surveys tend to understatetransport expenditures, while transport surveystend to overstate them.7 Subject to those caveats,it has been estimated that transport accounts forbetween 8 and 16 percent of household expen-ditures in a range of developing countries inAfrica.8 Estimates for major cities in some othercountries also fall in this range, with 15 percentfor an industrialized country such as France.

In the context of poverty assessment, the pro-portion of income spent on transport by differentincome groups is of more interest. Typically thereare two steps in transport expenditures corre-sponding to the progression from nonmotorizedto motorized public transport, and from publictransport to motorized private transport, respec-tively. Where those steps take place in any coun-try depends on income level and distribution, aswell as on the quality, availability, and cost of publictransport. Studies in Ouagadougou (Burkina Faso)and Dakar show that the highest quintile spends20 times as much on transport as the lowest quin-tile, but this only amounts to double the propor-tion of income (Godard and Olvera 2000). Invirtually all countries, richer groups spend a higherproportion of their incomes on transport than domost of those with lower incomes.

However, the proportion of income spent on trans-port varies greatly for the very poorest groups.

Some of the very poor may be forced to acceptprecarious living conditions in order to be able toaccess work. For example, a survey of pavementdwellers in Madras, India, showed that 59 percentwalked to work at no cost.9 In other circumstances,however, the burden of transport expenditure onpoor people may be very high. A study of low-income households in Temeke, Tanzania, 8 kilo-meters from the center of Dar es Salaam estimatedthat households spent between 10 and 30 percentof their incomes on transport, with an average of25 percent (Howe and Bryceson 2000). The upperlimit was very income constrained, while many low-income earners in the formal sector claimed thatthey could only afford public transport in theperiod immediately after being paid. Later, aftertheir pay was exhausted, they walked.

Given the high cost of transport, the time takenby the poor who are working to travel to workvaries greatly. The Madras pavement dwellers,walking less than one-half an hour to work, are apolar case of the tradeoff between transport costand residential quality. More generally, land-pricedifferentials reflect local environmental quality,and are likely to do so more as the middle classesgrow and environmental expectations rise. Evenin the largest cities, there may be areas of barelyhabitable or accessible land, such as those of the“favellas” (squatter developments) in Braziliancities, which are relatively close to areas of poten-tial employment but which are unserved by formaltransport providers.

The other polar case in the tradeoff concernsthose who live remotely in order to inhabit afford-able space, and who thus incur both high travelcosts and long travel times. As a result ofapartheid policies, the average distance of theblack townships from the central business dis-tricts (CBDs) of the seven largest South Africancities is 28 kilometers.10 Some poor people inLatin American cities—such as Lima (Peru) andRio de Janeiro—are also driven out to inexpen-sive dwelling space in remote locations, some 30or 40 kilometers out of the employment center(the average commuting time per day for the

URBAN TRANSPORT AND POVERTY REDUCTION 27

poorest group in Rio de Janeiro exceeds threehours). Such peripheral locations typically involveexclusion from a whole range of urban facilities,a deprivation only partly overcome by family orneighborhood solidarity.11

The transport patterns of poor people thus exhibita complex tradeoff among residential location,travel distance, and travel mode, in an attemptto minimize the social exclusion associated withlow earning potential. Differences in land pricesin developing countries generally reflect varia-tions in accessibility to the CBD or other centersof employment. Since good transport contributesto accessibility, it tends to drive up land rents anddrive out poorer residents, who can only affordto live closer in as pavement dwellers or in slumswhich are often inaccessible to motorized trans-port and are very difficult to inhabit.

The role of transport in this complex concept ofexclusion may be characterized as follows. The“income poor” make fewer trips, and more oftheir trips are undertaken on foot. For most pur-poses they are restricted to whatever services(usually poor services) that can be accessed withinwalking distance, making them “accessibilitypoor.” The journey to work may be relatively long.Even if it is not, it will use slow modes and maybe very time-consuming, so they are also “timepoor.” For poor people, and particularly forwomen, children, and the elderly, trip making isoften deterred because of their vulnerability aspedestrians, both to traffic accidents and to per-sonal violence, making them “safety poor.”Finally, there is evidence that long walking dis-tances and times also creates tiredness and bore-dom that reduces their productivity by addingan “energy-poverty” dimension to their depri-vation. In assessing transport provisions for poorpeople, it is therefore necessary to look at thetotal package that defines “exclusion,” and notsimply to look at the proportion of income, oreven of time, spent on transport.

Where public transport is not available, accessto a private mode of mechanized transport may

play a critical role in the extent of exclusion. Inthe United Kingdom, experiments with inexpen-sive car loans for rural workers who can onlyaccess jobs by private transport are improvingthe lot of some relatively poor people. The equiv-alent in poorer countries may be the develop-ment of mechanisms for inexpensive finance ofprivate bicycles—together with public investmentin infrastructure for the safe movement of thosebicycles. This is discussed further in chapter 9.

In addition to household characteristics, there arealso some specific personal characteristics thataccentuate deprivation. In most countries, over10 percent of the population has some form ofphysical disability imposing serious disadvantageboth in terms of mobility and safety (Merilainenand Helaakoski 2001). For the physically impaired,as well as for the elderly, public transport acces-sibility is often very poor and pedestrian facilitiesare often nonexistent or are blocked by parkedcars. Increasing attention is now being paid tothese groups in industrialized countries, andguides to good design practice are available.12

While some aids to mobility are expensive, andraise issues of expenditure priorities in circum-stances where affordability of basic transport itselfis an issue, many are not. Provision of pavementramps to make road crossing easier for wheel-chairs, tactile strips on station platforms to assistthe blind, large brightly colored signage to helpthe partially sighted, and well-designed grab barsand handles to assist the less mobile are all mat-ters of a more inclusive focus in design rather thanof expense.13 Good practice can be found indeveloping as well as in industrialized countries.14

Gender-related disadvantage is also endemic.Many activities typically undertaken by women(childcare, household management, informalsector employment, and so on) require them tomake more frequent and shorter trips than arerequired of men. They make more trips at off-peakhours and more trips that are off the main routes,and engage in more complicated multileg trips,all of which tend to make their movements rela-tively expensive for public transport to provide,

28 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

and hence more highly priced or more poorly sup-plied.15 Women are very vulnerable to these costcharacteristics because they frequently have lesscapacity to pay than do male household mem-bers, who, in many cultures, also control any bicy-cles or other vehicles available to the household.Cultural factors may constrain women’s abilitiesto use public transport or bicycles. In many coun-tries there is also a problem of the “social safety orsecurity” of public transport for women, especiallyafter dark.16 This may force them to depend onmore expensive alternatives. Peripheral locationmay be particularly damaging to women’s employ-ment potential.17 To confirm this, a heavy agendaof necessary gender-related research is required.This includes a need for more activity-based, asopposed to trip-based, research; better estimatesof the economic value of women’s time; and directevaluation of the impacts of some gender-relatedprojects.

Reforms aimed at improving economic efficiencymay sometimes have the immediate effect ofreducing employment of the poor or the rela-tively poor. Constraints on the development orbehavior of the informal transport sector, dis-cussed in chapter 7, may take away the onlysource of livelihood for some of the very poor.Rail reform has also often been associated withsubstantial severance of redundant staff, as hasoccurred in Buenos Aires. In World Bank projectsthis adverse side effect is mitigated by the impo-sition of resettlement provisions based on a policyof no detriment. But there is a wider issue. Notall impacts are so directly apparent. Identificationof distributional effects of infrastructure worksand of reform policies, and the fuller involvementof project-affected persons in decisions, is thus asine qua non for the avoidance of incidentaldamage to the interests of poor people.

Some general conclusions may be derived imme-diately from the analysis of the travel patterns ofpoor people.

a. “Exclusion” is multidimensional, so low travelcosts may be achieved through the accept-

ance of other heavy transport quantity, time,or quality penalties, or through the accept-ance of very bad housing conditions.

b. The transport capability of a household iscritically dependent on its stock of privatevehicles (bicycles, motorbikes, cars, and soon), as well as on its income and locationalcharacteristics.

c. The structure of provision of formal publictransport services tends to reflect and accen-tuate the distribution of poverty rather thanto compensate for it.

d. Some specific categories of people—defined in terms of age, gender, or infir-mity—may suffer particular disadvantage intransport terms.

PRO-POOR ECONOMIC GROWTHAND POVERTY REDUCTION

At the individual level, the urban poor are veryconscious that access to employment is critical totheir fight against poverty, and that the availabil-ity of good transport infrastructure and services isa basis on which this access can be achieved. “Thelack of basic road, transportation, and water infra-structure is seen as a defining characteristic ofpoverty,”18 but the relationship between urbantransport infrastructure and poverty reduction iscomplex. The “income poor” may in fact havechosen to live in poorly served peripheral loca-tions precisely because they are the places wheretheir overall welfare (in terms of availability of shel-ter, access to activities, and so on) is best served.High transport cost is then a symptom of theirpoverty rather than its fundamental cause. Hencetransport policies that improve the general eco-nomic viability of the city are very important topoor people. For example, the lot of poor peoplein Cairo, Egypt, has been improved more throughrelocation of their residences in order to improvetheir access to transport links, which are not pri-marily designed for poverty alleviation, thanthrough poverty-targeted transport investments.19

URBAN TRANSPORT AND POVERTY REDUCTION 29

This finding has a parallel at the macroeconomiclevel. World Bank research indicates that incomeof the poorest quintile of the population varies indirect proportion to national income (Dollar andKraay 2001). Moreover, there is no evidence of alag between increases of overall incomes and theincomes of poor people to suggest that benefitsaccrue to poor people only in a prolonged processof “trickle down.”20 That being so, urban transportinterventions that are particularly effective in gen-erating growth may also be particularly effectivein raising the incomes of poor people. Moreover,aggregate-level analysis of poverty and growthindicates that much previous public social sectorexpenditure has been poorly targeted, having littledemonstrable effect on either growth or distribu-tion, while, in contrast, policies to improve marketfunctioning has yielded proportionate benefits topoor people. Policies that most benefit the poorappear to be those associated with reducing gov-ernment expenditures and stabilizing inflation.

While these general analyses do not refer specifi-cally to urban transport, they highlight some criticalquestions about poverty-oriented urban transportinterventions. Many governments view urban publictransport policy as an instrument of their social policy.It is thus important both to establish how effectivelyurban transport infrastructure, service planning, andinvestment targets the needs of poor people, andto establish the indirect effect of urban transportpricing and financing policies on the poor throughthose policies’ impact on government expendituresand macroeconomic stabilization.

FOCUSING INFRASTRUCTUREPOLICIES

The selection and design of infrastructure invest-ments, whether in facilities for motorized or non-motorized road traffic or for rail traffic, mustconsider their impacts on the poor.

ROAD INVESTMENTSMost urban transport is road based. The avail-ability of an adequate road infrastructure is there-

fore a prerequisite for efficient urban movement.Some of the most intransigent urban transportproblems arise where the space devoted tomovement is both inadequate and poorly struc-tured; an example of this is Bangkok. Rapidlyexpanding towns need adequate road capacity,which may involve investment in limited-accessprimary roads as a structuring element. Some ofthe developing-country cities that appear to havethe best public transport facilities are also notablefor well-designed and managed road infrastruc-ture; an example of this is Curitiba.

But, there is a fine balance to be struck. Unlessroad space is already very abundant, there is adanger that more roads will simply encouragepeople to make extra trips to an extent that nul-lifies the intended reduction of congestion,increases auto dependence, and contributes tourban sprawl. In the absence of a strategic visionof the desired transport system, which addressesthe management of available space as well asthe planning of additions to it, and particularlyin the absence of efficient congestion pricing,piecemeal adjustment to emerging bottleneckswill almost certainly benefit the wealthy at theexpense of the poor (see chapter 6).

This finding has some consequences for the eco-nomic appraisal of urban road projects. As weargued in chapter 2, allowance should be madefor the effects of generated traffic in limiting theextent to which congestion can be reduced andtime and operating cost savings achieved. Thishas long been accepted in the context of moresophisticated, model-based appraisals in indus-trialized countries, but it tends to be forgottenin simpler, more abbreviated appraisals. Theproblems associated with differences in the waysthe rich and poor value time can also be handledby assigning a common value to all nonworkingtime for evaluation purposes.

Some more difficult issues remain unresolved.Because conventional transport planning is drivenby the willingness to pay (either demonstrated,in the case of commercial services, or synthesized

30 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

from behavioral studies, in the case of public infra-structure), relatively low value tends to be assignedto investments that cater to more dispersed andoff-peak transport needs. These needs ofteninclude those of the very poor, and of women.Moreover, much of the travel of poor people ison foot and typically receives low priority in con-ventional transport planning, which is often ori-ented to vehicle movement rather than to personmovement. If conventional evaluation cannot rec-ognize such categories of movement, it shouldnot be relied upon. Even participatory planningmethods may fail to accommodate this if theyunderrepresent both women and the very poor.

To some extent investments in road infrastructurecan be focused to specifically benefit poor people.Several such types of investment have been com-monly favored in Bank projects (see table 3.1).

a. Road investment and rehabilitation expen-ditures can be concentrated on major publictransportation routes so that public trans-port can benefit, as is the case in the currentKyrgyz Urban Transport Project.

b. Investments can be made in the provision orsegregation of routes for nonmotorized trans-port (NMT), including walking, to make NMTquicker and safer; this has been the case in recentprojects in Lima, Accra (Ghana), and others.

c. Road and sidewalk design can be more sen-sitive to the needs of disabled persons.

d. Road expenditures more generally may bedirected specifically to improve access topoor areas, or informally settled areas (forexample, the “pueblos jovenes” of Lima).

e. Particularly in informally settled areas (suchas in the Hanna Nassif project in Dar esSalaam),21 the use of employment-intensivemethods may be an important source ofincome for the very poor and may also createa local sense of ownership conducive togood maintenance.

NONMOTORIZED TRANSPORTThe most obvious policy mismatch is thatbetween the significance of NMT (walking andcycling) to poor people and the attention givento these modes, both in infrastructure design andin management. In the poorer countries morethan one-half of all trips are undertaken on foot,yet are typically treated as a peripheral issuerather than as a core element. Cycling offers arelatively inexpensive means of improving theaccessibility of poor people, but is often vulner-able both to accidents and crime; planning andproviding for cycling is often poor. Because thisis such an important issue, we devote a separatechapter (chapter 9) to NMT.

TABLE 3.1 POVERTY-FOCUSED URBAN TRANSPORT INTERVENTIONS: ROAD INFRASTRUCTURE

Source: Authors.

Specific Nature of Cost and Implementation Bank intervention impact fiscal impacts ease examples Maintaining public Faster and less- Moderate Easy Kyrgyztransport routes expensive public

transportPaving poor areas Access for public Moderate Easy Lima

transportBicycle and Safer trips; Moderate Moderate Lima, Accrapedestrian tracks encouraging NMTSeparation of NMT Safety; speed for Dhaka; on existing roads all modes

Low DifficultGuangzhou, China

URBAN TRANSPORT AND POVERTY REDUCTION 31

RAIL INVESTMENTS The poverty impact of investment in mass rapidtransit (MRT) has been very controversial. It mayimprove quality of the transport that poor peoplereceive and give them a wider choice of house-hold location, employment, and lifestyle. It canincrease their incomes indirectly through its effecton the efficiency of the urban system. But it mayconcurrently increase the fares that they mustpay. Prescriptions to subsidize MRT to offset theeffect of higher fares may actually militate againstthe interests of poor people unless it is clear thatthe subsidies are well targeted to them, do notprecipitate declines in service quality (which aremore harmful than higher fares would be), do notleak away through inefficient operation, and donot impose a burden on city finances by pre-empting other socially desirable expenditures.

The most basic consideration is what modes ofpublic transport poor people actually use. Insome cities in East Asia, buses are the transportof poor people and rail transport the mode ofthe relatively affluent. The same is not true inmany Latin American cities, where the averageincome of rail users is much below the averageincome, and very similar to that of bus users (table3.2). The pattern disclosed is one of great vari-ety, defying simple norms and emphasizing theimportance of relating policies sensibly to objec-tives on a case-by-case basis.

Metros may incidentally serve low-income areas,as in Cairo, but have rarely been designed specif-ically for that purpose. Even where they aredesigned to serve low-income areas, the opera-tions of the land market mechanism may resultin the benefits passing on to others. Insofar as ametro reduces travel time to central areas of thecity, it will tend to increase city-center land valuesand hence land rents at the newly advantagedlocations. Poor people only capture those ben-efits if they own the land themselves, and henceacquire the windfall capital gain, or are protectedagainst charges for the increased value of theland in property rents. That can be done if publichousing programs and mass transit developmentsare undertaken jointly by a development author-ity with a specific responsibility for the welfare ofpoor people, as has been done systematically inSingapore and more sporadically in a number ofother cities, such as Fortaleza, Brazil.

The link between efficiency and equity is some-times very subtle. For example, restructuring ofbus services to feed into higher-capacity trunklinks (either rail or bus) is commonly advocatedas a central part of integrated urban transportdevelopments, as in Singapore and Curitiba.However, this restructuring will tend to increasethe number of multileg trips involving separatepayment, which, given the typical flat or very shal-lowly tapered fare structure, can increase total

Note: MMS = monthly minimum salary.Source: World Bank, project files.

TABLE 3.2 INCOME OF USERS OF RAIL SYSTEMS IN BRAZIL

Percentage of City/region average Users’ average users below city

Rail system income (MMS) income (MMS) average income

Recife suburban rail (1997) 4.1 2.7 55.7

São Paulo Metro (1997) 13.6 13.8 57.4

São Paulo train (1997) 13.6 8.8 80.4

Rio de Janeiro suburban rail (1996) 10.0 3.2 85.5

Belo Horizonte (1995) 8.1 4.6 55.6

32 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

fare costs, particularly for those (often poorpeople) living in locations most remote from theMRT line.

That impact may be reduced by the introductionof multimodal through-ticketing systems, whichhave been shown to yield high benefits to usersin a number of countries (table 3.3). These multi-modal systems may be difficult to establish wherethere are a number of independently operatedmodes. Certainly it will tend to be easier to achievewhen the bus industry is relatively highly concen-trated (as in many Brazilian cities). Even where itis achieved, however, adverse distribution effectson the very poor may occur if the effect of incor-porating a high cost–high fare metro in a revenuepool is to raise fares, even for those who do notbenefit from the new investment. The lesson isthat wherever integration is introduced, it is nec-essary to analyze and design fare structures andcross-modal revenue support very carefully.22

Poor people may lose their jobs because of areform; both severance compensation andretraining finance is necessary to counteract thisconsequence. The poor may also suffer involun-tary displacement in the process of urban rail (orroad) infrastructure development, especiallywhere they occupy land illegally, or where theydo not hold title (that is, squatters). The expansion

of capacity of the suburban rail system in Mumbaiis estimated to involve the resettlement of 60,000persons. In that case full consultation with localslumdwellers’ associations at the project prepa-ration stage, as well as adequate financing fortheir resettlement, has been essential for avoid-ing harm to some of the very poor people.

PUBLIC TRANSPORT SERVICEPLANNING FOR THE URBAN POOR

Even in some of the most highly motorized citiesin Latin America, the average income of thosewho use cars is more than double the income ofthose who do not use cars (table 3.4). For the verypoor, transport service is synonymous either withNMT (mostly walking) or public transport, oftenvery inexpensively provided by the informalsector. Hence a poverty-oriented urban transportstrategy needs to concentrate on the movementof people rather than of vehicles.

That does not mean that municipal authoritiesshould be disinterested in the phenomenon ofroad congestion, which affects the movement offreight as well as people and which reduces theefficiency of the city. Congestion also tends todisadvantage those in crowded public transportvehicles even more than those in private cars. But

Source: Authors.

TABLE 3.3 POVERTY IMPACTS OF INTERVENTIONS: RAIL INFRASTRUCTURE

AND MANAGEMENT

Specific Nature of Cost/fiscal Implementation Bank intervention impact impacts ease examples Concessioning Improved service to Cost saving Moderate Rio de Janeiro,

users: fare effects Buenos Airesuncertain

Severance payments Protects (poorer) Small Moderate Buenos Airesworkers

Resettlement Protects disturbed Small/medium Difficult Mumbaiarrangements residents from

consequences of development

Converting suburban Improves speed and Moderate Moderate Fortaleza, Brazilrailways frequency

URBAN TRANSPORT AND POVERTY REDUCTION 33

it should be recognized that, even in relativelypoor cities, scarce but freely provided urban roadspace is increasingly appropriated by private carscarrying a small proportion of total “person-movements” made by the wealthy. This inequityimplies that priority should be given in the useof scarce road space to facilitating the movementof NMT and the more space-efficient public trans-port modes, rather than to private cars.

THE GENERAL SAFETY NET APPROACHFrom the conventional public transport supply side,the challenge of meeting the transport needs ofpoor people can be approached in two quite dif-ferent ways. In most socialist economies, publictransport was traditionally viewed as a basic socialservice. Even in some mixed economies, such asFrance and francophone Africa, concentration onthe concept of exclusion of identifiable spatialgroups has led to an emphasis on mobility as a“merit good,” a minimum supply of which is viewedas a social imperative. This approach leads to a“network completeness and integrity” approachto transport supply, with extensive fare reductionsor exemptions for disadvantaged groups. Itdepends on broad political acceptance of highlevels of subsidy of public transport operations, aswell as on state contributions, particularly to financecapital. The employment of a single private oper-ator, either under a management contract or undersome form of system concession, tends to gener-

ate a continued willingness to supply unprofitablelocations, both on the part of the operators, whodo not wish to abandon any part of their monop-oly domain, and on the part of the public authori-ties, who wish to sustain urban integration. But thisis probably not the least expensive or most effi-cient way of providing a basic network of services,as we will show in chapter 7.

Maintaining the basic social network is not alwayseasy. Even in France, as car ownership increasedand public transport patronage declined, therehave been pressures to reduce the fiscal burdenthrough tariff increases. In most francophonedeveloping countries, the attempt to maintainthe social obligations in the absence of a fiscalbasis for support resulted, initially, in the retreatof the traditional supply agencies to be suppli-ers only of those with fare concessions (who maynot be the very poor, but who may fall into cate-gories such as middle-class scholars) and ulti-mately in the bankruptcy and disappearance ofthe traditional supply.23

The safety net approach thus needs a secure finan-cial basis that is often lacking, given general budgetweaknesses of many developing-country cities.There are, however, two possibilities, discussed inmore detail later, that may avoid the need for anycharge on the general budget. The first possibil-ity is that, within a system of competitively tendered

a. Monthly income in dollars of 1995 for 1995 (Exchange rate is $1 = 1,000 pesos). Figure assumes 160 hours worked permonth. Source: JICA-Chodai 1996.

b. Monthly income in dollars of 1994. Figure is for 1997. Source: Centro de Estudios del Transporte del AreaMetropolitana 1999.

c. Monthly income in dollars of 1999 for 1999 (Peruvian currency seems overvalued). Source: APOYO 1999.Source: Authors.

TABLE 3.4 AVERAGE INCOME OF USERS BY TRANSPORT MODE

Average income Percentage of all Average income Average income car users/ motorized trips

City car users noncar users noncar users done by car

Bogotáa 462.4 196.8 2.3 19.2%

Buenos Airesb 607.2 299.1 2.0 40.0%

Limac 1,157.0 312.0 3.7 20.0%

34 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

franchises, the profits from the more profitableroutes may be used to support unprofitable serv-ices (chapter 7). The second possibility is that roadpricing may be used as a revenue source for a mul-timodal urban transport fund (chapter 10).

THE TARGETED SUBSIDY APPROACHThe alternative approach is to treat transportsupply more as a commercial business and totarget subsidies explicitly at disadvantagedgroups on a personal basis. In the UnitedKingdom, where supply of bus services is entirelyby the private sector in competitive markets,scholars and pensioners often benefit from farereductions or exemptions directly funded by therelevant line agencies; this funding hence appearsas commercial revenues to the operators. Transferof responsibility for “social” subsidies from theaccounts of the transport operators to those ofthe relevant line agencies is also being widelyadvocated as a means of addressing the declineof public transport service in many countries ofthe former Soviet Union.

The obvious advantage of this approach is thatthe fiscal burden on the community is lower. Italso has the merit of giving clear signals andincentives to the operators to adjust their serv-ices and fares in such a way as to maintain theirequipment in operation. The disadvantage is thatthere is no clear institutional channel throughwhich the more strategic and structural consid-erations concerning the role of public transportin urban development strategy, and the responseto the various externalities that impinge on urbanpublic transport, can be addressed.

COMPETITION, PRIVATIZATION, ANDPOOR PEOPLEIntroduction of competitive tendering of franchisesin major cities in Western Europe reduced costsper vehicle kilometer by up to 40 percent in realterms, and allowed higher service frequencies tobe maintained within constrained budgets thanunder traditional monopoly supply mechanisms.These advantages, which were first exploited inOrganisation for Economic Co-operation and

Development (OECD) countries, are now beingseen to be effective in developing economies(urban rail services in Buenos Aires, Argentina) andtransitional economies (bus services in secondarycities in Uzbekistan).

There are, nevertheless, some concerns aboutcompetitive private sector supply. These includethe loss of internal cross-subsidy, the abandon-ment of socially desirable services, and theincrease in fares associated with commercializa-tion. As discussed in more detail in chapter 6, allof these perceived problems can be overcomeby good design and administration of the com-petitive regime. The capability to combine somecentral service coordination with competitivesupply varies from country to country accordingto administrative capability and integrity, so thatthere is no single pattern that fits all economies.The lesson is that attention to the potential forcompetitive process can be a powerful contrib-utor to improvement of the services on whichpoor people depend. Some examples are pro-vided in table 3.5.

FARE POLICIES, SUBSIDIES, ANDBUDGET CONSTRAINTS

Controlling public transport fares, ostensibly tohelp the poor, may adversely affect service qualityunless supported by subsidy. This raises questionsboth about the concept of an affordable fare andabout the financing and targeting of subsidies.

THE CONCEPT OF THE AFFORDABLE FAREThe price and quality of service provided is obvi-ously important in assessing the impact of urbantransport on poor people. The concept of “afford-ability” of public transport is popular and seduc-tive, and governments frequently control publictransport fares because fares above some thresh-old level would be unacceptably burdensome topoor people. Although it was not originallyintended as a pricing policy prescription, the“Armstrong-Wright maxim” (that situations in

URBAN TRANSPORT AND POVERTY REDUCTION 35

which more than 10 percent of households spendmore than 15 percent of household incomes onwork journeys can be regarded as discriminatory)has often been interpreted as a reasonable rulefor determining the level of a politically admin-istered price.24

Caution should be advised over this maxim. First,the impact of any particular level of transport costson the aggregate welfare level of the householddoes not depend only on household income andthe price of transport. If shelter and heating areprovided very inexpensively through tax-financedpublic subsidies, then the proportion of dispos-able income for transport might be correspond-ingly higher. More important, though, price is notthe only thing that matters. There is evidence fromsocial surveys of public transport users inUzbekistan, the Kyrgyz Republic, and variousBrazilian cities that even relatively poor peoplemay be willing to pay more for the better serviceoffered by the informal sector, with their smallvehicles, compared with the inexpensive but slowand unreliable service offered by the public sector,with their traditional buses. Similarly, where metrofares are higher than those of buses, poor peoplemay choose to use the metro because of thebetter quality of service that it provides; this is thecase, for instance, in Cairo.25

The most serious problem, however, is that manygovernments control general fare levels without

making any accompanying fiscal provision forsubsidies. The rationale for this, often explicitlystated, is that it will force operators to cross-sub-sidize unprofitable services from profitable serv-ices, leading to cross-subsidy of poor people bythe rich. In practice, in many countries there is nosuch basis for cross-subsidy (the rich do not usepublic transport and there are no profitable serv-ices from which to squeeze cross-subsidy finance).In these circumstances the main effect is toreduce the quality, and eventually the quantity,of public transport service.

Sometimes the adverse effect of the failure ofthe traditional formal sector is attenuated by thedevelopment of informal sector services, usuallyprovided with smaller and less expensive vehi-cles, frequently at fares above the controlledformal sector fares. Insofar as this informal sectorservice involves the provision of a quality of serv-ice that could be improved by the traditionallarger vehicle if allowed to operate at the fareadopted by the informal sector, it represents adistortion with unintentionally adverse effects onpoor people. The lesson is that tinkering with thesymptoms of poverty may actually make thingsworse for the poor. The policy prescription aris-ing from this is that the likely supply outcomesof different levels of fare intervention and sub-sidy should always be estimated. Fares shouldbe set at levels consistent with the outcome pre-ferred by poor people as shown by surveys, and

Source: Authors.

TABLE 3.5 A POVERTY-FOCUSED AGENDA: PUBLIC TRANSPORT SERVICE PLANNING

Specific Nature of Possible cost/ Implementation Bank intervention impact fiscal impact ease examples

Introduce competition Cost-reduction service Cost saving Moderate Uzbekistan and in public transport growth Kazakhstan

Public transport Faster, safer trips Medium Moderate Pusan, Republic interchange of Korea; Manila

Bus priorities Faster, less expensive Low Politically difficult Bangkoktrips

Develop informal sector Lower-cost service None Moderate Uzbekistan, secondary cities

36 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

not on the basis of some “normative” conceptof what an affordable fare might be.

FINANCING TARGETED SUBSIDIES Many countries have extensive lists of categoriesof passengers qualifying for free or reduced-faretravel. Rarely is there any specific mechanism forremunerating suppliers for these fare exemptionsor reductions. This has two effects. First, it meansthat some passengers are paying more, or receiv-ing poorer service, than would otherwise be thecase in attempts to secure cross-subsidy. Becausethe rich often do not use public transport, thismeans, at best, subsidy of the poor by the poor.Second, it creates a vested interest of benefitingnontransport agencies (health, education, police,and so on) in maintaining a subsidy for their par-ticular user group, that they might not favor if ithad to be financed from their own budgets. Thelesson is that, in the interests of poor people, anypublic transport fare reductions or exemptionsshould be carefully considered in the light ofother uses that might be made of the resourcesinvolved. That consideration is probably bestensured by putting the responsibility for financeof fare exemptions or reductions directly on thebenefiting line agencies, with the obligations onthe transport operators contingent on the receiptof the appropriate compensation.

Even where a fiscal basis for corrective action toreduce poverty exists, the question arises as towhether intervention in the transport sector isthe most appropriate use of such funds. Theanswer to that question turns partly on the rela-tive efficiency with which funding can be targetedin different sectors, and partly on the politicalfeasibility of taking poverty-reducing actions invarious sectors. The relationship between theaverage income of the users of specific modesand the overall average incomes is usually known,or can be established, and the distribution ofgeneral fare subsidies can thus be assessed.There are two main impediments to the use ofgeneral subsidy, however. First, there may bewide variations of income among users of a spe-cific mode, so that targeting is very imprecise.

Second, there is substantial evidence that a largeproportion of subsidy to public transport throughdeficit financing of public sector monopoly oper-ators “leaks” away, either through inefficiency ofoperations or through the capture of the subsidyby organized labor in the supply industry. Whilethere remain some problems in achieving ade-quate targeting, it is clear that public transportusers include the poorest, and usually excludethe richest, groups, so that in the absence of anymeans of transfer through taxation structures, itmay be a reasonably good discriminator in manydeveloping countries. As further discussed inchapter 6, competition is the best protectionagainst leakage of the benefit to suppliers or theiremployees.

CONCLUSIONS: A STRATEGY FOR POVERTY-FOCUSED URBAN TRANSPORT

Inadequate and congested urban transport isdamaging to the city economy and harms bothrich and poor. But the simplistic solution ofincreasing road capacity in an attempt to speedup the movement of vehicles, accompanied bypublic provision of fare-controlled public trans-port, is likely to be inequitable (because it leadsto a progressive decline of public transport serv-ices) and ineffective (because it will tend to gen-erate more congesting car traffic). Rather, thereis need for a more poverty-focused policy (seetable 3.6) reflecting the following general con-clusions concerning the impacts of urban trans-port policies:

• Costs should be properly charged for allvehicle movements, both to secure efficientuse of infrastructure, and to generate asecure financial basis for urban transport pro-vision.

• The importance of walking and other NMTactivities, and the special needs of the mobil-ity-impaired should be recognized both ininfrastructure design and in traffic manage-ment.

URBAN TRANSPORT AND POVERTY REDUCTION 37

• Ill-judged policies on general public trans-port fare controls in the absence of securesubsidy mechanisms can actually harm poorpeople.

• Constraints on the informal transport sectoroften harm poor people. The message hereis that policies for the informal transportsector need to be framed with their impactson poor people carefully taken into account.

• Absence of competition in public transportis likely to both increase costs and reducesupply to poor people. A preference forstable, disciplined supply should not beinterpreted as a case for uncontestedmonopoly.

• Efforts to secure multimodal integrationneed to be carefully managed to ensure thatthese efforts do not increase the number oftimes poor people must pay per trip, andthat fares on the services on which they areparticularly dependent do not increase.

• Attention needs to be given to financing ofsupport mechanisms, avoiding deficit financ-ing of monopolist suppliers, and, whereverpossible, targeting very specific groups.

• Because of the effect of transport infra-structure investment and transport pricingpolicies on land values, it is important thatostensibly poverty-oriented urban transportinterventions be integrated in a broaderstrategy incorporating housing, health, edu-cation, and other social service policies.

Based on these principles, there is a rich agendaof urban transport policies that are both pro-growth and pro-poor, yet that are consistent withthe fiscal capabilities of even the relatively poor-est countries.

NOTES

1. World Bank 2000c.2. In this volume, poor countries are defined as

those falling below the threshold that qualifies forInternational Development Association borrowingterms, currently an average annual gross nationalproduct per capita of $885. Much of the distribu-tional analysis referring to the conditions of poorpeople uses data on the bottom quintile of incomeper capita within a country (and hence are not strictlycomparable across countries). Little transport-relateddata are linked to the commonly quoted absolutepoverty standard of $1 per day per capita.

3. Schwela and Zali 1999. 4. An early study in Salvador, Brazil, showed

that the lowest income groups on average madeone trip per person per day, compared with threetrips per day for the highest income groups, witha constant distance per trip of 6 kilometers acrossall income groups. See Thompson 1993.

5. The recently observed exceptions to this insome of the major cities of Latin America mayreflect the increasing insecurity of travel, whichis now affecting the nonpoor as well as the poor.

Source: Authors.

TABLE 3.6 POVERTY-FOCUSED URBAN TRANSPORT INTERVENTIONS: FINANCE STRATEGIES

Specific Nature of Cost/ Implementation Bank intervention impact fiscal impact ease examples

Subsidy finance Line agencies to Uncertain Moderate Russian reform finance exemptions; Federation

better focus of support

Public transport Enables use of Low Moderate São Paulo fare integration faster modes and Fortaleza,

Brazil

Congestion pricing Direct impact small. Generates Difficult Kuala Provides basis for revenue Lumpur, public transport Bangkok

improvement

38 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

6. A curious exception to this, which is worthfurther investigation, is the city of Ouagadougou,where public transport is almost nonexistent and57 percent of households falling in the bottomincome quintile possessed a motorized two-wheeler.

7. Godard and Diaz Olvera 2000.8. Godard and Diaz Olvera 2000.9. Madras Metropol i tan Development

Authority 1990.10. De St. Laurent 1998.11. Cusset 1998.12. European Conference of Ministers ofTransport 1999b.13. Merilainen and Helaakoski 2001.14. Wright 2001.15. Diaz Olvera, Plat, and Pochet 1998.16. Gomez 2000.

17. In a study of squatter resettlement in Delhiin the late 1970s, it was found that male employ-ment among the colony decreased by 5 percent,but female employment decreased by 27 per-cent after resettlement (Moser and Peake 1987).18. Narayan 2000a.19. CATRAM 2000.20. Dollar and Kraay 2001.21. Howe and Bryceson 2000.22. Recent MRT development packages in someBrazilian cities have been subject to detailed analy-sis of the effects of changes in bus service struc-tures and integrated ticketing arrangement onthe money, and generalized costs of transit forzones with populations at different income levels.23. Teurnier and Mandon-Adolehoume 1994.24. Armstrong-Wright 1986.25. See White 1999.

THE SCALE OF THE ISSUE

For many poor people in developing countries, theinability to access jobs, education, and health facil-ities is viewed as the most serious constraint on theirquality of life (Narayan 2000a). Motorized transportis often crucial to poor people, but because theyfrequently live and work in the shadow of motor-ized transport, they are also the most vulnerable toits adverse environmental impacts. Environmentalprotection must therefore be seen as an essentialpart of the task of improving the quality of life ofthe poor, and not as a luxury to be purchased atthe expense of the poor’s mobility.

A comprehensive strategy for transport and theurban environment should cover the built envi-ronment, including land take, urban form, visualintrusion of infrastructure and traffic, and culturalheritage; the social effects of transport in caus-ing occupational or locational resettlement andcommunity severance; personal safety and secu-rity; and the more commonly recognized prob-

lems of noise and air pollution (local, regional,and global). All of these concerns must be for-mally addressed in a typical project’s environ-mental assessment.

In this review we deal with matters concerningthe built environment in chapters 2 and 6, withthe social effects in chapter 3, and with safetyand security in chapter 5. In this chapter we dealonly with various forms of pollution, especiallyair pollution.

In developing countries an estimated 0.5 millionto 1 million people die prematurely each year asa result of respiratory and other illnesses causedby exposure to urban air pollution. This is a largernumber than those dying as a consequence ofurban traffic accidents, although because of theage distribution of those afflicted, traffic acci-dents probably reduce life-years more. Exposureto lead contributes to behavioral problems andlearning disabilities in urban children. It alsoreduces the quality of life.

TRANSPORT AND THE URBAN ENVIRONMENT 39

TRANSPORT AND THE URBAN ENVIRONMENT4Road transport contributes significantly to urban air pollution in many

countries. The World Health Organization estimates that suspended particulate

matter leads to the premature death of over 0.5 million people per year. The

economic costs of air pollution have been estimated to be equivalent to about

2 percent of gross domestic product in many countries. Incorporation of

environmental issues within an urban transport strategy requires the

identification of the main transport-generated pollutants (usually suspended

particulate matter, lead, and ozone) and the mobilization of technical, fiscal,

and system management controls on fuel and vehicle technology to reduce

these pollutants. Frequently these will also contribute to a desirable reduction

of greenhouse gas emissions.

40 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

GLOBAL WARMING

It is now generally agreed that a global climatechange is occurring. It also appears that thepoorer countries stand to suffer most as a con-sequence of this change, with estimated costs inthe range of 5 to 9 percent of gross domesticproduct (GDP) for some of the poorer countries—several times greater than the relative effect inindustrialized countries.1 It is estimated that thetransport sector is responsible for about 25 per-cent of emissions of the gases contributing toglobal warming in industrialized countries, butonly about one-half this amount in developing-country cities.2 While the proportion appears tohave been stabilized in the Organisation forEconomic Co-operation and Development(OECD) countries, it is still growing in the devel-oping countries as motorized transport increases.This increase in motorized transport is concen-trated in urban areas. Although controversy con-tinues over the optimal greenhouse gas (GHG)reduction strategy, and the distribution of actionbetween industrialized and developing countries,it is accepted that some mitigating strategy iscalled for in all countries.

Despite this, GHG mitigation has a negative con-notation in many developing countries, whereexhortations to limit GHG emissions are per-ceived as a denial of the right to develop the serv-ices and lifestyle being enjoyed by industrializedcountries. The apparent unwillingness of someindustrialized-country governments to take strongaction also does not help, while the inherent long-term and nonlocal nature of the negative impactsof GHGs feeds this attitude.

A “business-as-usual” scenario for the transportsector offers little prospect of relief. The princi-pal components determining the level of GHGemissions in transport are the level of activity (inton or passenger kilometers), the modes of trans-port used, the energy intensity of each mode,and the mix of fuels used. In the industrializedcountries, transport activity has increased paripassu with economic growth, the shift to the pri-

vate car and air transport has worsened themodal balance effect, and changes in the bal-ance of fuels has had little impact. It is only thedramatic increase in fuel efficiency that has actedas a brake on GHG emissions. Even so, trans-port GHG emissions continue to grow in theindustrialized countries even though overall emis-sions have been stabilized. Economic growth inthe developing countries similarly threatens todominate any attenuating effect of technologyimprovement.

To avert this outcome requires a combination oftransportation policy reforms in the short termand technological changes in the longer term.The question is how to get such policies adopted.The suggested key to changing this situation is tolink GHG-mitigation policy initiatives to goalsthat are perceived to be of immediate relevance(such as local air pollution and balance-of-pay-ments considerations) and to try to uncouple, orat least “flex,” the link between economic growthand GHG emissions from the transport sector.3

We start from the observed synergy betweenGHG reduction and local environmental and eco-nomic interests. The GHGs that most contributeto global warming in the transport sector includecarbon dioxide (CO2), methane, and nitrous oxide(N2O). Emissions of CO2 are directly proportionalto the quantity of carboniferous fuel consumed;other things being equal, reduced fuel con-sumption will reduce economic costs and globalpollution simultaneously. Better traffic flow con-ditions typically reduce fuel consumption per kilo-meter. In chapter 6 we advocate both road-basedtraffic management and traffic restraint measuresto that end, while the measures that are sug-gested to improve public transport in chapter 7and to improve nonmotorized transport (NMT)in chapter 8 should have similar effects. Moregenerally, local air quality improvement programsfor urban transport in middle-income countriessuch as Mexico and Chile have also shown somecollateral benefits for reducing GHG emissions.4

While the rest of this chapter focuses primarilyon local air pollution reduction, much of it also

TRANSPORT AND THE URBAN ENVIRONMENT 41

relates, by implication, to the global warmingissue. Because these types of interventions can beshown to be in the immediate self-interest of cityresidents themselves, it is believed that concen-trating on exploiting the synergy between GHGreductions and local economic and environmentalinterests is likely to be the most productive strate-gic stance.

This must be supported by economic incentives.Fossil fuel consumption is influenced directly byfuel choice, vehicle size, and fuel efficiency, andindirectly by individual and corporate decisionson activity location and style, and transport mode.Emissions of methane come largely from leak-ages of gasoline, diesel, and unburned naturalgas and are thus susceptible to influence by fuel-ing infrastructure improvement. The strongestincentive to fuel economy in actions in all thesedimensions is the monetary cost of fuel. It hasbeen shown that pollution controls supported byappropriate prices or taxes are much more effec-tive than is the use of regulations alone.5 In chap-ter 10 we discuss this in detail, arguing for fuelprices that at least cover the full social costs offuel consumption.

Unfortunately there is not always a synergybetween local air pollution and GHG emis-sion–mitigation measures. The current generationof diesel vehicles appears to be more damagingto public health than are gasoline or gas-poweredvehicles. Thus, while diesel is a particularly effi-cient fuel from the point of view of reducing GHGemissions, only the new generation of clean dieselsshould have a role in GHG strategy. Furthermore,mitigation measures for local pollution focus onemissions of vehicles in use, whereas the entirelife cycle (from well to tailpipe) is relevant for analy-sis of GHG emissions. Worldwide policies to refor-mulate transport fuels to mitigate local pollutionby means of severe hydrotreating (particularlyrecent moves in North America and the EuropeanUnion [EU] to limit sulfur in gasoline and diesel to10–50 parts per million by weight [wt ppm] orlower) make refinery processes increasingly energyintensive, increasing GHG emissions. N2O can

increase significantly when catalysts used to con-vert nitric oxide (NO) or nitrogen dioxide (NO2)begin to be deactivated.

A strategic response is therefore required toaddress situations where the synergy is weak ornegative and where a tradeoff between local andglobal effects appears. In this context, the GlobalEnvironment Fund (GEF), a multidonor fundadministered by the World Bank, has promotedthe concept of the “global overlay.” This is a pro-cedure in which measures developed to targetother objectives are subject to a scrutiny explor-ing the possibility and cost of modifying them toyield GHG reductions. In this way an attempt ismade to identify and concentrate on those areaswhere GHG reductions have the least “opportu-nity cost.” Through GEF’s Operational Program11, on transport, GEF funding is available bothfor the development of new globally friendly tech-nologies and, on an incremental cost basis, forother interventions that pilot promising currentGHG reduction policies. An early grant under thisscheme is supporting the development of infra-structure for bicycle movement in an outer areaof Manila. The Bank has also established a“Prototype Carbon Fund” to foster the interna-tional transfer of certified emission reductionsunder the Clean Development Mechanism, asdefined in Article 12 of the Kyoto Protocol.

Technological measures to secure GHG reduc-tion perform primarily through the replacementof the vehicle stock. This may take up to 20 yearsto complete for cars and up to 30 years for freightvehicles. In developing countries, however, mostof the change of vehicle stock is through growth,so that strategies affecting new vehicles may havea more rapid effect on emissions per unit of activ-ity. Hence, with GDP rising, the best way toachieve a decline in GHG emissions by transportvehicles is a combination of policy reforms in theshort term and technological changes in thelonger term. Because most vehicle users aredriven by economic motives, this implies a needfor strong support from taxation and pricinginstruments. “Closing the loop,” so that revenues

42 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

from increased taxation or charges for vehiclesor fuels are seen to contribute to improvementin the transport sector, is likely to be a necessarycondition to secure acceptance of the policies atthe political level. This is further discussed in chap-ter 10.

NOISE AND OTHERDISTURBANCES

Noise from transport appears to be consideredmuch less seriously in developing countries thanin high-income countries. While there have beenstudies of the physical damage resulting fromexposure to occupational noise, these have beenmostly in manufacturing establishments. Thelevels experienced in developing-country streets,while not pleasant, approach but do not exceedthe lower limits above which noise is consideredan occupational hazard. Attitude surveys do notshow urban transport noise to be perceived as aserious hazard.

Other disturbances exist. Heavy road traffic vol-umes can make roads dangerous and difficult tocross, causing community severance. It isreported that in Jakarta, businesspersons rou-tinely take taxis just to get safely to the other sideof the busiest thoroughfares. Barriers, foot-bridges, and tunnels may reduce the danger butincrease the severance, particularly in countrieswhere these facilities appear to be constructedand located to improve motor vehicle flow ratherthan to assist pedestrians (see chapter 5).

Severance can in some circumstances be reducedby the grade separation of motorized traffic andmass transport from pedestrian movements. Inestablished cities, tunneling is often too expen-sive (and technically difficult for major flows ofroad traffic for ventilation reasons) so that eleva-tion is the only viable possibility. This can causesignificant visual intrusion as well as restrict thedispersion of fumes from traffic remaining atground level. The elevated rail transport systemso dominates Silom and Sukhumvit roads in

Bangkok that the government has decided toavoid any further elevated mass transit in the cen-tral area. The problems of community severanceand visual intrusion can, to some extent, bereduced by good engineering design. The mainlesson from experience is that community sev-erance and visual intrusion have often developedbecause of inadequate coordination amongstrategic planning, transport infrastructure invest-ment planning and design, and management ofprivate participation in infrastructure. That empha-sizes some institutional requirements, to whichwe will return in chapter 11.

URBAN AIR POLLUTION

By far the greatest environmental concern abouturban transport in most cities relates to local airpollution. Designing an appropriate strategy, withlimited resources, to address this concern requirescareful identification of priorities, both in select-ing targets and in selecting instruments.6 Thissection first identifies the major local air pollu-tants produced by urban transport, and assessesthe significance of their contribution to the totalurban environmental burden. On this basis, wethen discuss four main types of instrument forconsideration as components of a strategy forreducing the impact of transport on the urbanenvironment. The first two types—actions ontransport vehicles and transport fuels—are pri-marily technological. The other two types ofinstrument—traffic management and fiscal instru-ments—while requiring technological supportfor effective implementation, are “softer” policyinstruments. In some countries, such as Mexico,public transport improvements are sought pri-marily for environmental reasons.

MAJOR TRANSPORT-GENERATED AIRPOLLUTANTS AND THEIR SIGNIFICANCE Vehicular emissions are very damaging to health.

a. High lead concentration in the bloodstreammay increase incidence of miscarriages in

TRANSPORT AND THE URBAN ENVIRONMENT 43

women, impair renal function, and increaseblood pressure. Most significantly, it retardsthe intellectual development of children andadversely affects their behavior. More leadis absorbed when dietary calcium intake islow, in cases of iron deficiency, when thestomach is empty, and by the young, so poormalnourished children are particularly sus-ceptible to lead poisoning. Suspended par-ticulate matter, particularly particles fromvehicle emissions and tire wear that fall pre-dominantly in the submicron range, are ableto penetrate deep into the respiratory tract,cause respiratory problems, exacerbateasthma, and damage lung function. Thereis also a growing consensus that dieselexhaust poses a serious cancer risk.7

b. Carbon monoxide (CO) inhibits the capac-ity of blood to carry oxygen to organs andtissues. People with chronic heart diseasemay experience chest pains when CO levelsare high; at very high levels, CO impairsvision, manual dexterity, and learning abil-ity, and can cause death.

c. Sulfur oxides (SOx) which are emitted indirect proportion to the amount of sulfurpresent in fuel, cause changes in lung func-tion in asthmatics and exacerbate respira-tory symptoms in sensitive individuals; theycontribute to acid rain and to the formationof secondary particulate matter.

d. Oxides of nitrogen (NOx) cause changes inlung function in asthmatics, contribute toacid rain and secondary particulate forma-tion, and are a precursor of ground-levelozone. Both diesel- and gasoline-fueled vehi-cles contribute to NOx emissions.

e. Ozone is responsible for photochemicalsmog and decreases pulmonary function inindividuals taking light to heavy exercise.NOx (which is emitted in significant quanti-ties by gasoline- and diesel-fueled vehicles)and photochemically reactive volatile organic

compounds (to which emissions by gaso-line-fueled vehicles contribute) are the twomain precursors of ozone.

While all of these emissions are potentially dam-aging, their incidence and their health impactsdiffer substantially, both between pollutantsa n d b e t w e e n re g i o n s . Wo r l d H e a l t hOrganization (WHO) studies of megacities,although now somewhat dated, show that,although health norms of all major pollutantsare widely exceeded, the significance of theproblem varies considerably (WHO 1992). Leadexcesses over norm are very serious whereleaded gasoline is used, but not usually else-where. Excess of CO is typically not nearly asgreat as that of fine particulate matter, partic-ularly in countries where the consumption ofgasoline is relatively low compared with that ofdiesel. Significantly elevated levels of ambientSO2 tend to come from the combustion of coalmuch more than from the transport sector.Ambient NO2 concentrations are often belowthe WHO guidelines, but are on the increase,as are those of ozone (figure 4.1).

When risk assessment of susceptibility to physi-cal excesses are combined with evidence ofhealth impacts from dose and response analysis,studies in a number of cities (for example,Bangkok; Cairo; Mexico City; Quito, Ecuador;and Santiago) have indicated that the greatestdamage to human health comes from exposureto fine particulate matter (particles smaller than2.5 microns in aerodynamic diameter, or PM2.5)and to lead. Depending on topographical andmeteorological conditions, ozone can also be aserious health problem in large metropolitanregions, as it is in Mexico City and Santiago.8

Transport is not the only source of urban air pol-lution. In particular, industrial and domestic useof fossil fuels—especially heavy fuel oil, biomass,and brown coal—is a significant source of ambi-ent particulate matter and sulfur dioxide (SO2),especially in temperate regions such as Chinaand Eastern Europe.

44 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

Any strategy to reduce pollution from the trans-port sector needs to be seen in this broader con-text.9 However, urban transport is generallyidentified as a high-priority action area for sev-eral reasons.

• Urban traffic is a large contributor to themost harmful fine particulate emissions, andit is responsible for up to 80 to 90 percentof atmospheric lead in cities where leadedgasoline is still used, for the greatest part of

FIGURE 4.1 OVERVIEW OF AMBIENT AIR QUALITY IN SELECTED CITIES: SUBJECTIVE

ASSESSMENT OF MONITORING DATA, VARIOUS YEARS IN THE 1990S

OECDLondonLos AngelesNew YorkTokyo

East AsiaSeoulBeijingJakartaBangkokManila

South AsiaKarachiMumbaiDelhi

Latin AmericaLimaMexico CitySão PauloBuenos AiresRio de Janeiro

Central Asia,Africa, andEurope

TehranCairoLagosMoscow

POLLUTANTSCO NOX Lead SPM SO2 O3CITIES

Low pollution Relevant WHO/national guidelines are normally met (short-term guidelines may beexceeded occasionally).

Moderate to heavy pollution Relevant WHO/national guidelines exceeded by up to a factor of two (short-term guidelines exceeded on a regular basis at certain locations).

Serious problem Relevant WHO/national guidelines exceeded by more than a factor of two.No data available Information on ambient polluted concentration not available.

CO = Carbon monoxide SPM = Suspended particulate matter O3 = OzoneNOx = Oxides of nitrogen SO2 = Sulphur dioxide

Source: The data for figure 4.1 have been assembled by Asif Faiz and Surhid Gautam, World Bank, from various sources.These include: World Bank project reports; Urban Air Quality Management Strategy reports, World Resources Institute; astudy of seven cities undertaken for the Latin America and the Caribbean Clean Air Initiative; Indian Central Pollution Boardreports; and unpublished data obtained by direct inquiry from various cities (São Paulo, Santiago, Bangkok).

TRANSPORT AND THE URBAN ENVIRONMENT 45

CO emissions, and for significant contribu-tion to the formation of ground-level ozone.

• Large stationary sources of air pollution,which are often located at a distance fromdensely populated city centers, disperse intothe higher layers of the atmosphere whilevehicles emit near ground level in highlypopulated areas. Consequently, vehiclescontribute more to human exposure thantheir share in total emissions loads wouldindicate. In a study of six megacities, vehi-cles accounted for only 6 percent of emis-sions in tons emitted but 32 percent ofaverage exposure for the population.

• The urban transport sector is one of veryrapid growth and change, which makes itvery susceptible to positive and to protec-tive actions. An understanding of the envi-ronmental significance of alternativetransport policies and actions may enablegrowing cities to avoid the environmentalimpacts already endured in the megacities.

FUEL POLICY Fuel policy measures may affect the pollutingcharacteristics of existing fuels, the selectionamong available fuels, and the total amount offuel consumed.

Improving fuel qualityFor the World Bank’s client countries, the first stepin improving the quality of transport fuels is tophase out lead in gasoline. Lead has historicallybeen added to gasoline as an octane enhancer.Because of its toxicity, there is now a worldwidemove to ban its use in gasoline. More than three-quarters of the gasoline sold worldwide today isunleaded. Virtually all OECD countries and manylarge developing and transition countries, includ-ing Bangladesh, Brazil, Honduras, Hungary,Malaysia, and Thailand, have already eliminatedlead in gasoline.10 Some very large countries,including Indonesia, Venezuela, and most coun-tries in Sub-Saharan Africa, remain to be converted.

In the absence of other significant sources of lead,eliminating lead additive in gasoline can reduce

ambient concentrations of lead to less than 0.2micrograms per cubic meter (mg/m3) and the levelof lead in blood to lower than 5 micrograms perdeciliter (mg/dl), below the 10 mg/dl level nowconsidered by many health organizations to bethe appropriate norm. The necessary strategyshould attend both to the introduction ofunleaded fuel and the elimination of leaded fuel.

Eliminating lead additive from gasoline can alsotrigger wider environmental improvement. Theavailability of unleaded gasoline throughout acountry is a prerequisite for the introduction ofcatalytic converters to reduce the emissions ofNOx, CO, and hydrocarbons. The level of COemissions can also be reduced by incorporatingoxygenates into gasoline. It is important, how-ever, to stress that lead elimination should notbe carried out in isolation because many of thegasoline-blending components used to increaseoctane after elimination of lead have their ownadverse health effects. Excessive presence ofbenzene and total aromatics in unleaded gaso-line would be of particular concern. However,controlling gasoline volatility and adjusting refin-ery operations and processing units can managethese blending components at a reasonable costduring the lead phase-out process.

Sulfur in diesel and gasoline generates emissionsof SO2, causes acid rain, and contributes to par-t iculate emissions. I t can be reduced byhydrotreating the base fuels. However, in coun-tries where the carbon component of vehicularparticulate matter remains high, it may not makeeconomic sense to attempt to match NorthAmerican and EU sulfur standards immediatelyin order to mitigate particulate emissions fromdiesel engines.11 Rather, tightening of standardsshould be carefully phased in the light of coun-try-specific circumstances. In some extreme cases,proposed emissions standards are incompatiblewith the transport fuels available on the market.For example, insistence on Euro 2–compliantbuses when the sulfur level in the diesel availablein the country may be as high as 5,000 wt ppm isnot technically coherent. Some regional effort

46 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

may also be necessary to harmonize refining andimport standards to avoid local black markets inhigh-sulfur diesel. Transport, environment, andenergy policies must always be carefully aligned.

It is not enough to regulate fuel quality. In manydeveloping countries, old, poorly maintained vehi-cles dominate vehicle fleets. This reduces the costeffectiveness of imposing stringent fuel specifi-cations.12 Moreover, in some countries transportfuels are routinely adulterated. For example, theaddition of (lower-cost) kerosene to gasoline inAsia (box 4.1), cross-contamination of diesel withcrude oil, and addition of lead additives to gaso-line downstream of refineries or terminals inCentral Asia and the Caucasus all increase vehicleemissions. Regular fuel-quality monitoring,together with costly penalties for noncompliance,could help enforce fuel standards more effectively,although preventing local adulteration is likely toremain very difficult as long as there is any finan-cial incentive to engage in the practice.

Fuel-quality requirements are location specific,depending on climatic conditions, ambient con-centrations, vehicle fleet characteristics, and soon. For example, Chile and Mexico have morestringent standards than do other Latin Americancountries because of the particular pollution char-

acteristics of their capital cities. Countries thathave domestic refineries merit special attentionin this context because refinery processes areintegrated, and changing the specifications ofone fuel can affect the quality of other fuels andoverall refinery economics. While many countrieshave fuel standards, most of these standards arestipulated in the form of fuel composition. In theUnited States, the combination of allowing theoption of fuel certification on the basis of vehi-cle emissions, in lieu of fuel composition, andregional differentiation of specifications hasallowed the refining sector the freedom to seekflexible least-cost solutions for meeting specificvehicle emissions standards.

The governments own many refineries in devel-oping countries. Some are not operated eco-nomically at present. Revamping refineries toimprove fuel mix and quality is likely to renderthem even less commercially viable. Under thesecircumstances the government may resist requir-ing changes in fuel quality, or will embrace themonly while maintaining import protection throughrestrictions or high tariffs. In some developingcountries, the net cost to society of improving fuelquality by importing superior fuels would be lowerthan the costs resulting from the use of domesti-cally manufactured fuels with less-stringent spec-

BOX 4.1 FUEL ADULTERATION IN THAILAND

Adulteration of heavily taxed gasoline by highly subsidized kerosene was a serious problem inThailand in the early 1980s. The government introduced a number of measures, including:

• Dyeing the kerosene blue• Requiring kerosene to be sold in 20-liter containers only• Extensive enforcement efforts by the police.

Although these measures had some effect, sales of kerosene remained high until oil taxes wererestructured in 1986 and the tax on kerosene increased between 1986 and 1991 to remove theincentive to adulterate. However, the incentive to adulterate gasoline with untaxed industrial sol-vents remains, and such adulteration is a continuing problem.

Source: Jitendra Shah, private correspondence.

TRANSPORT AND THE URBAN ENVIRONMENT 47

ifications. Downstream petroleum–sector reformthrough transfer of ownership from the govern-ment to the private sector, coupled with liberal-ization of petroleum product trade and theintroduction of competition, can therefore result inimproved fuel and, ultimately, urban air quality.

Substituting cleaner fuelsA range of alternative fuels considered to becleaner than conventional hydrocarbons continueto be under investigation or development inindustrialized countries. For substitutes to beattractive in the developing world, they must beseen not only as addressing locally perceivedenvironmental problems but also as economi-cally viable at the individual and national levels.It is in that context that the potential of new fuelsfor urban transport in the developing world mustbe appraised.

Compressed natural gas (CNG) is a relativelyclean fuel.13 Natural gas is available in abundancein many developing countries that do not haveother indigenous fuel resources—such asArgentina, Bangladesh, and Thailand—and henceis potentially of great balance-of-payments sig-nificance. As transport use of CNG alone is notenough to justify the development of gas fieldsand the construction of gas transport and distri-bution infrastructure, the availability of CNG fortransport is closely linked to its availability throughcity gas distribution networks. These exist in manylarge cities in Bangladesh, Brazil, Colombia,Indonesia, Pakistan, Eastern Europe, and theformer Soviet Union. Some governments havealready specifically mandated the use of naturalgas as the transport fuel in highly polluted areas(for example, for taxis in Buenos Aires and, morerecently, for buses and all pre-1990 auto-rickshawsand taxis in New Delhi, India).

The environmental benefit of CNG is not undis-puted. Although recent tests by the New YorkCity Transit Authority have shown that natural gasbuses have better local pollutant emission char-acteristics than do ordinary diesel buses, theiremission advantages are significantly lowered or

even eliminated when compared with diesel-powered buses running on ultra low sulfur dieseland equipped with a continuously regeneratingparticulate trap. Similarly, although the overallglobal warming impact of cars fueled by CNG islower than that with gasoline (especially wherethe natural gas would otherwise be flared), gaso-line vehicles that are converted to natural gassuffer from potentially high leakage of the GHGmethane. Although some good retrofitted kits(for example, those used in Argentina) work effi-ciently, many do not. For example, the conver-sion of a fleet of Mercedes buses in Rio de Janeirois reported to have increased most emissionsexcept those of particulates.

A second disadvantage concerns the economicand technical sustainability of the technology indeveloping countries. The New York City TransitAuthority reported that CNG vehicles pay a 30to 35 percent energy premium over diesel vehi-cles. Technically, of a fleet of 40 CNG buses inJakarta, 20 were out of operation in mid-2001due to maintenance problems. Dual-fuel vehi-cles also carry some extra cost penalties and reli-ability problems. Again in Jakarta, a fleet ofNissan dual-fuel vehicles is now being run assingle-fuel (diesel) vehicles for technical reasons.

The economics of CNG are complex. Becausethe choice of fuel generally rests with the busi-ness or individual, the critical factor is ultimatelythe cost and convenience of CNG compared withthat of other fuels. As far as the fuel is concerned,the Intergovernmental Panel on Climate Change(IPCC) estimated the wellhead gasoline equiva-lent production cost of CNG to lie between 70and 90 percent that of gasoline or diesel, so thatgiven differences in distribution and storagecosts, the cost at the pump (excluding taxes)could be very similar. The actual real resourcecost thus depends critically on the local avail-ability of fuel and density of the distribution net-work.14 As far as vehicles are concerned, there isextra cost associated with the CNG engine (orits conversion), the fuel control system, and thefuel tanks. Together these increase the cost of a

48 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

basic vehicle, whether a bus or a car, by up to 30percent.15 The convenience factor is also impor-tant because CNG vehicles lose significantamounts of luggage and passenger space to fueltanks; at low penetration levels, refueling caninvolve some dead-running and can be time-con-suming; vehicle range may be reduced by over50 percent, doubling the refueling frequency.Considering all these factors, international evi-dence suggests that, except for some very heavymileage vehicles, the pump price differencewould need to be about 50 percent of the pro-duction cost of liquid fuels for natural gas to beattractive to users. Hence natural gas would seemto require strong fiscal encouragement if it is tobe more than a niche fuel.

Liquefied petroleum gas (LPG) is a mixture oflight hydrocarbons, mainly propane and butanes.It is easier to distribute and store than CNG,16

and although the octane number of LPG is notas high as that of natural gas, it has excellent anti-knock characteristics allowing dedicated propanevehicles to take advantage of engines with slightlyhigher compression ratios than can be used withgasoline. The limited amount of highly reactivehydrocarbons and the low sulfur content of LPGin comparison with gasoline or diesel are some ofLPG’s good environmental features; it does, how-ever, contain olefins, which are photochemicallyreactive. LPG-powered three-wheelers are com-mercially available in Bangkok, and have alreadyeffectively replaced the old two-stroke gasoline-powered “tuk-tuks.”

The main problems in introducing LPG to thetransport sector are the supply sources and dis-tribution system. Several countries already importsignificant amounts of LPG. India, Pakistan, andSri Lanka, for example, import about 30 to 40 per-cent of their LPG demand. On the distributionside, LPG is stored under pressure both insidethe vehicle and in the refueling tanks. Specialrefueling equipment is needed to transfer thepressurized liquid from storage tanks to the vehi-cle and to ensure that no LPG escapes duringrefueling. As with CNG, the required investments

in LPG distribution and refueling stations havenot been made in most developing countries,and the need for such stations remains a con-straint on widespread LPG use.

Ethanol and methanol. True biofuels (that is, thosewithout substantial fossil fuel use hidden in har-vesting and processing) would give a real reduc-tion in GHG emissions, but these are still elusiveat costs competitive with those of gasoline or diesel.The only long-term effort to promote biofuels fortransport in developing countries, the sugar-ethanolprogram of Brazil, appeared attractive as a meansof saving foreign exchange when oil prices wereat their peak, but has lost most of its attractiveness;the new car market in Brazil is now almost exclu-sively for gasoline vehicles. In any case, manyexperts argue that only alcohol produced from cel-lulose can truly claim a GHG benefit.

Electric vehicles. Electric road vehicles are quietand nonpolluting at their point of use and haveobvious attractions as urban vehicles, whetherpowered directly, as in the case of electric trainsor trolleybuses, or indirectly, as in the case ofsome buses, small vans, and cars. A program forelectric three-wheelers is being undertaken inKathmandu, Nepal (box 4.2). While these havethe lowest environmental impacts at point of use,their overall environmental impact depends onthe way in which electricity is generated andstored, and the disposal problems associatedwith expired batteries, which can be substantial.Whatever that environmental balance, the marketattractiveness of electric-powered transportdepends, as is the case for other fuels, on its eco-nomic attractiveness in terms of overall cost andconvenience. At present the economics of elec-tric vehicles are far from favorable.17

Battery technology is central to the economic suc-cess of battery electric vehicles. Lead-acid bat-teries, currently used in electric vehicles, take 6 to10 hours to slow-charge, emit hydrogen whenrecharging (requiring indoor recharging facilities tobe well ventilated) and still have very limited range.Other battery types are still in the development

TRANSPORT AND THE URBAN ENVIRONMENT 49

stage, and significant efforts are being directedto electric–internal combustion engine hybridsrather than pure electric engine vehicles. The eco-nomics of electric vehicles also depends on theprice of electricity. The power sector in many devel-oping countries is currently undergoing reformand restructuring. The long-term viability of elec-tric vehicles should be evaluated from the stand-point of market-based power pricing. Given thecurrent state of technology, electric vehicles wouldnot be expected to have widespread applications;with carefully considered government interven-tion, however, they could play a useful, thoughlimited, role in extremely polluted traffic corridors.Moreover, the greenhouse characteristics of elec-tric vehicles depend critically on a full fuel cycleanalysis; if the electricity is produced from pres-ent mixes (coal, oil, or gas), there may be no green-house benefit at all compared with a small internalcombustion engine.

Hybrid diesel-electric vehicles are now beingdeveloped with some success and tested inindustrialized countries under a GEF grant. Theircost is similar to that of a heavy CNG vehicle.Onboard diesel engines operated at constant

load to maintain battery power for peak demandsand in sensitive areas can give a 30 percentenergy savings compared with a conventionaldiesel vehicle.

Hydrogen fuel cell. This is a widely advocated“sunrise” technology, at least for heavy-dutyurban vehicles. Fuel-cell buses are already beingused in trial projects, including a program fundedby the United Nations Development Programme.The environmental performance of these vehi-cles—depending on the source of the hydro-gen—can far exceed that of CNG or improveddiesel engines, and their life-cycle operating costsare projected to be lower than that of CNG ordiesel. It is possible that we will see such vehi-cles deployed in active service in urban fleets(buses and urban freight delivery applications) inthe industrialized countries in about a decade. Itis unlikely, however, that they will have early appli-cation in the developing countries.

Influencing fuel consumptionIn industrialized countries technological improve-ments in engine efficiency tended to decreaseboth local and global pollution in the 1970s and

BOX 4.2 ELECTRIC THREE-WHEELERS IN KATHMANDU

In 1993, faced with growing air pollution in Kathmandu, the government of Nepal banned thesale of new diesel three-wheelers imported from India. In the same year, the Kathmandu munic-ipality invited the United States–based Global Resources Institute to design an electric three-wheeler for the city. The pilot vehicle, known as “safa tempo,” was put on the road in September1993, and U.S. Agency for International Development (USAID) support expanded to buy eightvehicles, promote the program, and train mechanics. When USAID funding ended in 1996, two localbusiness groups started assembling, servicing, and operating electric vehicles, which sell for about$6,000. In September 1999 the government finally banned all existing diesel three-wheelers. Thereare now six suppliers of electric vehicles, and as of May 2000, the city has over 500 battery-pow-ered three-wheelers, each capable of carrying up to 12 passengers. This is the world’s largest fleetof electric passenger transport vehicles for use on public roads. The future is not secure, how-ever. In May 2000 the government approved the import of 300 15-seat vans with the same pref-erential import duties accorded to the electric vehicles. In May 2001 it decided to ban new ortransferred registration of all three-wheelers, including safa tempos, although that decision waslater rescinded under pressure from donors.

Source: The Wall Street Journal Europe, May 30, 2000.

50 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

the first half of the 1980s, although effectivereductions in fuel consumption have subse-quently been lost in the United States byincreases in vehicle size. Possibilities for furtherimprovement in new-vehicle fuel economy arisefrom reduction of vehicle size or weight; directfuel injection, lean-burn technology; measuresto increase the share of diesel (which may, how-ever, adversely affect local pollution); and opti-mized engine transmission systems and hybridvehicles. Fuel-consumption standards, such asthe corporate average fuel efficiency (CAFE) stan-dards imposed by the U.S. government and thosemore recently agreed on between Associationdes Constructeurs Européens d’Automobiles(ACEA—the organization of the European motorindustry) and the European Commission, haveattempted to force the pace of technologicalprogress in this respect.18

In developing countries, fuel economy is affectedby what is occurring in the industrialized coun-tries through the availability and cost of importedsecond-hand vehicles, but they also have specialproblems of their own. Fuel economy is often lowbecause of poor vehicle maintenance, fuel adul-teration, and a number of other factors. Usinggasoline with an octane number that is lower thanthat recommended by vehicle manufacturers—either because lower-octane gasoline (for exam-ple, 80 research octane number [RON]) is availableand is less expensive or because gasoline is adul-terated with kerosene—can decrease fuel econ-omy, lead to knocking and ultimately to enginedamage, and to higher emissions. Hence a ratherdifferent focus may be appropriate for policy onfuel economy in developing countries.

The scope for improving fuel economy is great-est in countries where a large fraction of vehicleshave low engine compression ratios. In the newindependent states of the former Soviet Union,many vehicles run on gasoline with a motoroctane number as low as 72. Increasing theoctane and the engine compression ratio willresult in fuel savings—and reductions in GHGemissions—in the long run.

A staged strategy for fuelsThe timing of the development of fuel-cell tech-nology may have a significant part to play instrategic thinking about the alternatives. CNGrequires significant investment in gas production,distribution infrastructure, and vehicles, whichwould only be recouped over a substantialperiod. The incremental cost of cleaning up theperformance of diesel is significantly lower. Exceptin a few cities where the CNG infrastructure isalready in place, it may make more sense, as aninterim strategy, to concentrate on getting thediesel vehicles currently on the road performingwell and cleanly, to gradually reduce sulfur con-tent in diesel fuel, and to bring in cleaner diesel-vehicle technology in new vehicles on the road,while waiting for the cost of fuel cells to becomeviable.19

VEHICLE POLICYEnvironmental issues for vehicles can be dividedinto those relating to improving new-vehicle tech-nology, those relating to replacing the existingstock by more environmentally friendly technol-ogy, particularly with respect to motorcycles, andthose relating to the use of the existing vehiclestock. Although we discuss vehicles and fuelsseparately here, in practice they must be con-sidered simultaneously.

Improving vehicle technologyLargely in response to stricter vehicle emissionsrequirements, considerable progress has alreadybeen made in developing vehicle hardware toreduce emissions. Efficiently operated three-waycatalytic converters can reduce exhaust CO andhydrocarbon emissions of gasoline vehicles byas much as 95 percent and NOx by over 75 per-cent. Similarly, state-of-the-art diesel vehicles thatuse ultra low sulfur diesel fuel and continuouslyregenerating traps can be almost as clean as vehi-cles that use CNG. Such advanced technologiesare unlikely to offer cost-effective solutions invery low income countries in the near future, buttheir existence makes the point—an importantone to keep in mind in formulating incentive poli-cies—that not all diesel vehicles are alike.

TRANSPORT AND THE URBAN ENVIRONMENT 51

The EU, Japan, and the United States lead theworld in setting stringent vehicle emissions stan-dards and fuel specifications. These countries arepursuing the best available technology for fur-ther reducing emissions from new vehicles. Thecontrol measures include a combination of thefollowing: dramatic reduction of sulfur in gaso-line and diesel, to extend the useful life of thecatalyst and to enable new catalyst technologiesfor reducing NOx

20 and particulate emissions;new measures for control of tailpipe emissions(for example, particulate traps with regenerationfor diesel engines); emerging vehicle technolo-gies such as common-rail direct-injection dieselengines; and the use of alternative fuels for verylow emission or zero emission vehicles (as aremandated in California). Although the rest of theworld will probably adopt these standards andtechnologies some day, the issue for developingcountries is how to phase in these measures cost-effectively.

In encouraging the use of catalytic convertersin developing countries, a number of conditionsneed to be satisfied to ensure that they func-tion effectively. These conditions are (a) wideavailability of unleaded gasoline and, prefer-ably, complete phase-out of leaded gasoline,to eliminate the chances of misfueling; (b) dif-ferentiated taxation during the transition periodto prevent misfueling;21 (c) a reasonably low levelof sulfur in gasoline, preferably lower than 500wt ppm; (d) specification of the emissions per-formance levels and the length of time duringwhich the catalyst system must meet thoselevels; and (e) effective inspection and mainte-nance (I/M) to ensure that converters are oper-ating properly. If these conditions cannot befully satisfied, the additional cost associatedwith the installation of converters may not bejustified by the benefits.22 Even where effectiveuse of catalytic converters is considered feasi-ble, governments should consider specifyingemissions levels for new vehicles rather thanmandat ing cata ly t i c conver ters per se .Retrofitting in-use vehicles with catalytic con-verters is not usually considered cost effective.

Options for motorcyclesMotorcycles account for about one-half of thevehicle fleet in many Asian cities (and up to 75percent in some). In many cities motorcycles offersubstantially greater speed and flexibility ofmovement than do inadequate and congestion-bound bus services, while being broadly com-parable to them in cost. In Taiwan, China, theownership is already 0.55 per capita. The major-ity of motorized two- and three-wheelers in Asiaare powered by two-stroke engines, which arepreferred because they have a (now only slightly)lower capital cost than do four-stroke engines,have more power (higher power-to-weight ratio,higher specific output, higher torque, and lowrevolutions per minute response) for a given dis-placement, and are simpler to self-maintain.

Unfortunately, conventional two-stroke enginesare environmentally very damaging because oftheir inherent combustion technology; their poormaintenance and misfiring, particularly at coldstart; and their frequent and excessive use oflubricants not manufactured for use in two-stroke-engine vehicles.23 In Delhi, India, for example,45 percent of particulate emissions and two-thirdsof unburned hydrocarbon emissions in the trans-port sector are estimated to come from two- andthree-wheelers powered by two-stroke engines.Despite this, no country has issued standards forPM (particulate matter) emissions for two- andthree-wheelers, largely because there is noproven methodology for measuring oil droplets(although smoke standards might be used as aproxy standard for two-stroke engines).

The balance of advantage between two-strokeand four-stroke technologies is changing. Thedifference in capital cost is rapidly beingeroded.24 Taking into account differences in fueleconomy and the lower cost of crankcase oil thantwo-stroke oil, the total annual operating cost ofowning and operating a four-stroke can alreadybe lower than that of owning and operating atwo-stroke. Furthermore, the decision of one ofthe major market economies, Taiwan, China, toimpose stringent environmental standards on

52 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

two-stroke motorcycles starting in 2004 is likelyto shift the economies of scale in production infavor of the four-stroke, and to concentrate futuretechnical development on the four-stroke tech-nology. Already, Japanese manufacturers appearto be concentrating their sales efforts on four-strokes, even where two-strokes are still legal.

Basic two-stroke technology can be improved atsource by emissions-reduction technologiesincluding electronic control for fuel metering andimproved scavenging characteristics, or by after-treatment.25 Two-stroke engines manufacturedin India, meeting year 2000 emissions standards,emit very little PM. Unfortunately, the durability ofcatalyst and secondary air systems is limitedbecause of the high concentration of hydrocar-bons in the exhaust gas.26 Retrofitting ofimproved technology is relatively expensive andineffective for this category of small vehicle. It istherefore advisable that governments specifyemissions levels for new vehicles and adopt poli-cies to secure premature replacement for highpolluters, rather than mandate catalytic convert-ers per se. The key is to switch to four-strokeengines.

New technology does not deal with the problemof the high average age and low replacement rateof vehicles currently in use. For these vehicles,emissions on the road are usually much higherthan are emissions levels found during tests, dueto bad operational and maintenance practice.They are estimated to emit more than 10 timesthe amount of fine particulate matter per vehiclekilometer than is a modern car, and only a littleless than a light diesel truck.27 In some countries,such as Bangladesh, low-octane gasoline is used(80 RON, compared with a recommended mini-mum of 87 RON), and gasoline is often adulter-ated with inexpensive kerosene, causing enginemalfunction. It is also common for the proportionof lubricant in the fuel mix to be two to three timesthe recommended level, even though this bothincreases cost and reduces performance.Nevertheless, given the strong current preferencefor the use of two-stroke engines, mandatory

scrappage of existing fleets is likely to be politi-cally difficult. Phasing out more gradually might beachieved by the imposition of differential licensefees, according to technology and location, toencourage more-polluting vehicles to move intorural areas where pollutant concentration is less.The same effect might be achieved by refusingto license high emitters in the cities or limitingthe maximum permissible age of vehicles. Subsidyof environmentally advantageous scrap-and-replace programs might also help. For two-strokevehicles remaining in use, the first step might beto mandate an appropriate lubricant standard.Maintaining the recommended mixtures (2 per-cent oil for two-wheelers and 3 percent for three-wheelers) might be facilitated by the sale ofpremixed fuel. Driver education is an importantsource of improvement in this respect (box 4.3).Improved I/M are also powerful instruments.28

The evidence thus suggests a number of strands ina strategy for addressing the motorcycle problem.

• The information basis for policy needs to beimproved through more systematic meas-urement of ambient air quality, the intro-duction of PM standards for new two-strokes,public information campaigns on fuel mixand vehicle maintenance, identification oflocally cost-efficient retrofit technologies,and the establishment of secure commoninstitutional commitment (police, inspectionagencies, and so on).

• Standards on new motorcycles should beregularly reviewed and progressively tight-ened in the light of technological possibili-ties. This process would be enhanced ifgovernment announced long-term objec-tives with respect to fuel quality and vehicleemissions standards.

• Existing two-stroke performance should beimproved by introducing minimum gasolineoctane requirements (87 RON), mandatingappropriate oil standards, introducing pre-mixed fuel, introducing systematic I/M, andoffering incentives for conversion to cleanfuels for three-wheelers.

TRANSPORT AND THE URBAN ENVIRONMENT 53

• High polluters should be eliminated througha program including the statistical identifi-cation of high-polluting ages and categories,testing followed by improvement to stan-dard or buy-in of nonconforming vehicles,the introduction of tax- or license fee–incen-tives for low polluters in cities, and the reduc-tion of import duties on environmentallybenign vehicles and parts.

Using the existing stock of vehiclesThe share of emissions is not uniformly distrib-uted over the vehicle fleet. A fraction of all vehi-cles—ill maintained, often old—is typicallyresponsible for a disproportionately high amountof pollution from the transport sector. If these“high emitters” (typically, commercial vehiclesand public transport vehicles, including, in someplaces, two- and three-wheeler taxis with two-stroke engines) can be repaired or eliminatedpermanently, a considerable reduction in pollu-tion can be achieved at relatively small cost.

The implementation of such a scheme is far fromsimple. To be cost effective, any scheme target-ing high emitters should identify polluting vehi-cles with high annual vehicle kilometers traveledoperating in densely populated areas. Old vehi-cles in very poor condition may be candidates

for retirement. Those that are highly pollutingbut are better maintained may be considered forrepair or for retrofitting with more recent vehicletechnology. Complicating the issue is the factthat in some cities, such as Cairo, the proportionof old, gross polluters may be very high, with theresult that targeting does not leave very manyvehicles out of the scope.

APPLYING ENVIRONMENTAL STANDARDS To apply environmental standards, it will be nec-essary to design appropriate vehicle inspectionand maintenance programs, scrappage pro-grams, as well as ensure that related policies areappropriate, and provide the correct incentives,including those related to domestic taxation,trade liberalization, and public expenditures.

Inspection and maintenance programsVehicle emissions standards and technologiesare not effective without proper maintenance.Poorly maintained vehicles are high emitters andare responsible for a disproportionate share oftotal vehicle emissions. Data collected in India inNovember to December 1999 during a series ofI/M “clinics” for two-wheelers indicated thatminor vehicle repairs improved fuel economy byan average of 17 percent and reduced CO emis-

BOX 4.3 THE BENEFITS OF INFORMATION: LUBRICATING TWO-STROKES

It is a commonly believed fallacy in some countries that if a little oil is a good thing, more must bebetter. In Dhaka, Bangladesh, many operators of two-stroke three-wheelers use less-expensivestraight mineral oil rather than the recommended two-stroke oil. Because this gives less lubrica-tion, they also tend to use as much as 8 or even 12 percent mix compared with the recommended3 percent for two-stroke oil. Since two-stroke oil cost twice that of mineral oil in June 2000, followingthe recommended procedures would actually give a savings of 25 to 40 percent in oil costs aswell as reduce engine damage and emissions. But bad habits are inbred in a generation of oper-ators and mechanics. To counter such misinformation, training sessions for mechanics and “autoclinics” for three-wheeler taxi drivers were therefore held in 2000 under the Energy SectorManagement Assistance Program. Similarly, in Pakistan, pamphlets containing basic informationon fuel and lubricants have been distributed by the Hydrocarbon Development Institute.

Source: Masami Kojima, World Bank.

54 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

sions by 44 percent. A well-run, uncorrupt I/Mprogram should be able to strengthen theenforcement of emissions standards significantly.

Introducing effective I/M programs has proveddifficult. For example, recent experience inWuhan, China, showed that in roadside testing,93 percent of vehicles fell short of the standardeven though 97 percent of those tested in thesame period at the I/M station met it.29 Thus,wider use of spot-checking equipment capableof identifying major problems from any relevantvehicle type, and of fixed I/M stations for morethorough examination and follow-up of vehiclesso identified, holds promise of significant impacton the pollution problem. An appropriate systemof fines, and controls thereof, could also make

such system self-financing. As far as off-road test-ing is concerned, experience from various partsof the world suggests that an I/M system basedon centralized, high-volume, inspection-only cen-ters with computerized emissions measurementto minimize tampering and corruption, such asthe I/M system in Mexico City, is likely to be moreeffective than is a decentralized system with alarge number of private garages participating inthe I/M. If the proper controls are in place, theprivate sector can be an important partner inoperating effective I/M programs (box 4.4).

The usual reason for not properly maintainingvehicles is to avoid out-of-pocket expenses. Thereare, however, certain maintenance practices thatwould actually yield cost savings. An example is

BOX 4.4 PRIVATE SECTOR VEHICLE INSPECTION AND MAINTENANCE IN MEXICO CITY

Mandatory testing for vehicle emissions in Mexico City was introduced in 1988. Initially testingwas done in government test-only centers as well as in private garages that were permitted toboth test and repair. Although the private sector in Mexico undertook testing more economically,initially as many as 50 percent of vehicles were estimated to obtain passes falsely. A more limitednumber of private test centers were therefore subsequently licensed for testing only. The Mexicanexperience shows that an effective testing system must evaluate emissions levels accurately, andissue and enforce certificates without corruption. To achieve this it is necessary that:

• The legal framework provides sanctions to be applied for failure to carry out the testing pro-tocols correctly

• The testing stations must be subject to monitoring by independent bodies, and sanctionsmust be properly applied

• Repair work should be separate from testing• The pass certificate must be easy to monitor • There should be sufficient monitors (for example, traffic police) to ensure a low probability of

evasion by vehicle owners• The fine for not displaying a legal emissions test certificate should be sufficient to discour-

age evasion• The technology of testing should exclude the possibility of temporary “tuning” to pass the

test• The number of licensed centers should not be too large, to avoid garages being “soft” to

increase market share • All testing centers should be subject to rigorous implementation of protocols and inspec-

tion of their procedures.

Source: World Bank 2001.

TRANSPORT AND THE URBAN ENVIRONMENT 55

use of the correct kind and amount of lubricant inthe two-stroke engines that are common in SouthAsia (box 4.3 above). At a minimum, wide publiceducation campaigns should be undertaken topromote cost-effective practices.

Scrappage programsWhen emissions standards are enforced effec-tively, the cost of owning old vehicles actuallyincreases, making vehicle renewal more attrac-tive.30 Vehicle retirement and scrappage pro-grams can further encourage this phenomenonas long as it is possible to identify gross emittersthat are operating at a high annual rate of vehi-cle kilometers and that still have reasonableremaining economic lives. Considerable care isneeded in designing such schemes, however.Using age as a proxy for high emissions does notalways identify the worst cases. If the program isdirected at urban pollution, it must also preventthe import of old vehicles from outside the cityto take advantage of the retirement bonus.Evidence in Europe that may be relevant for someof the higher-income developing countries sug-gests that—because gross emitters are typicallyowned by low-income households that are oftenin no position to purchase much newer vehicles—cash-for-scrappage schemes may be more effec-tive than cash-for-replacement schemes.31 In themost polluted areas, it may even be possible touse cash-for-scrappage schemes to influencemodal split.31

In many developing countries, particularly thosein which particulate emissions are the most seri-ous pollution concern, commercial vehicles(buses, trucks, and taxis) are the greatest con-tributors to urban air pollution. Some countrieshave been quite successful in stimulating earlyreplacement of these vehicles. For example,Hungary offered large cash incentives for replace-ment of old buses and trucks with new vehiclescomplying with most recent emissions standards($3,600 in 1997 dollars for a bus replacement). Inthe early 1990s Chile combined tax incentiveswith preferential treatment of environmentallybenign vehicles in competitively tendered fran-

chising arrangements to remove the most-pol-luting diesel buses from the urban transport fleet.

Domestic taxation policiesThe importance of taxing fuels at a level thatreflects the externalities, as well as the bordercosts and any proxy charges for road mainte-nance costs, was emphasized in a recent WorldBank sector policy paper.33 Although the cost ofmitigating impacts is logically a better founda-tion for defining appropriate levels of environ-mental taxation valuation of damage, theevidence on marginal damage costs of the sixmegacities, quoted above, suggests taxes around60 percent of the import cost of gasoline and 200percent of that of diesel fuel.

These estimates emphasize not only the impor-tance of the absolute levels of fuel taxes andprices but also their relative levels. Around theworld the retail price of diesel is typically lowerthan that of gasoline, because of differential tax-ation. The trend is particularly pronounced inSouth Asia. In Bangladesh, for example, the retailprice of gasoline was almost double the price ofdiesel in 1999. The price differential, togetherwith the low profit margin fixed by the govern-ment for the sale of gasoline, has led to the adul-teration of gasoline with kerosene and, as anunintended consequence, higher particulateemissions from vehicles. In Pakistan, where theprice difference is even higher, the diesel-gaso-line ratio consumed in the transport reached 5.3to 1 in fiscal 1999/2000, which is very high by inter-national standards. Relative tax rates may thusbe encouraging the highest-mileage urban vehi-cles (taxis, mini-buses, and so on) to switch fromgasoline—not to CNG or other clean fuels butto diesel, the fuel with the most damaging urbanenvironmental impacts. A large price differentialbetween kerosene and gasoline, based on theuse of kerosene by the poor as a domestic heat-ing fuel, leads to illegal addition of kerosene togasoline, resulting in higher pollutant emissions.

The usual reason for low tax rates on diesel is itsuse for heavy interurban freight movement and

56 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

agricultural purposes. Given that the healthimpact of particulate emissions is likely to belower in low-density interurban and rural areasthan in cities, the emphasis on the economicrather than the environmental impact of dieselfueling outside urban areas may not be unrea-sonable. Hence it is necessary to develop taxstructures that protect the urban environmentbut that do not discourage use of the most eco-nomical fuel by agriculture or intercity freightvehicles. One possible way of doing that is toidentify the most damaging vehicle types (cars,urban mini-buses, and vans) and use high dutieson those types of diesel vehicle, rather than fueltaxation, as the means of changing the balanceof economic advantage. That is already beingdone in some countries by exempting clean vehi-cles from import duties or vehicle license duties.Another option is to increase the tax on dieselto make the price of automotive diesel compa-rable to that of gasoline, but to rebate industrialand agricultural users of diesel. This approachhas been adopted successfully in Chile. Anotheris to tax automotive diesel more and use a dyeto distinguish automotive diesel from diesel forother uses.

Setting relative tax levels is a complex issue. Inprinciple, one should identify the emissions valuesof different pollutants and structure vehicle andfuel taxes to reflect differences in the summedvalue of emissions for different vehicle types. Inpractice, however, emissions levels depend notonly on the fuel type and composition but alsoon where and how it is burned. Moreover, the evi-dence on the health costs of different pollutantsremains sketchy. At best, then, the use of fiscalincentives would be somewhat rough and ready.Nevertheless, on the basis of successful applica-tion of fiscal incentives in industrialized countries,it is possible to confidently recommend differen-tiation of excise taxes between leaded andunleaded gasoline, and between diesel and CNG.

Tax structures that discourage the purchase ofnew vehicles—for example, registration fees orexcise taxes based on the market value of the

vehicle—may have adverse environmentalimpacts that should be weighed against theirperceived distribution effects. Possible measuresinclude replacing import tariffs on new vehicles byownership taxes reflecting the environmentalquality of the vehicle, eliminating the systemunder which vehicle registration fees are pro-portional to the book value of the vehicle (whichmakes it more expensive to own new vehiclesthan old ones), and minimizing the tax on thepurchase of new vehicles. The fiscal impact ofthese changes might need to be offset by anincrease in direct charges for road use. In somemiddle-income countries, it may be necessary toconsider a safety net to offset the immediateimpact of making it more costly for the poor tooperate old vehicles, particularly in the taxi trade.

Trade liberalizationIt is common for developing countries to use tar-iffs or trade barriers to protect domestic indus-try and to prevent the expenditure of scarceforeign exchange on luxury goods that are notessential to economic growth. Where either ofthese arguments is applied to the import of vehi-cles, the effect is likely to be the protection ofoutdated technology. Liberalization of vehicletrade is hence an important step, particularly incountries that have automobile-manufacturingfacilities. The removal of barriers that hinderaccess to the technology available in the rest ofthe world would enable consumers in these coun-tries to meet tighter emissions standards at leastcost. Rules such as local content requirements(for example, requiring that 70 percent of thevehicle weight or content must be produceddomestically) often result in inefficiency and inheavier vehicles if the percentage is based onweight. High import tariffs on new vehicles, rigidlicensing schemes for imports, and quotas are alllikely to slow the rate of vehicle renewal, withpotentially adverse impacts on air quality (pollu-tion); their distributional impact must thereforebe weighed against their environmental effects.

Free trade in used cars raises the question of“environmental dumping.” Among the largest

TRANSPORT AND THE URBAN ENVIRONMENT 57

recipient markets for used cars are Cyprus,Jamaica, Peru, Sri Lanka, and the RussianFederation, while Japan remains the largest iden-tifiable single source of used-car exports. Theforces driving the export of used cars by indus-trialized countries will become stronger as emis-sions standards become more stringent andregulations concerning the end-of-life vehicle areimplemented, as in the EU.

In the interest of environmental protection, thegovernment may limit the age of the vehiclesthat may be imported. For example, Hungaryset the age limit to 10 years old in 1991 and pro-gressively reduced it to 8, 6, and finally in 1997,4 years old. Chile bans imports of used vehiclesaltogether. However, the purchasing pattern ofvehicle owners should be carefully balancedagainst the hypothetical environmental advan-tage of restricting the import of old vehicles. Ifcommercial operators (and, in some of the tran-sition economies, low-income households) arein no position to buy relatively new vehicles, suchan import restriction constrains the supply, andincreases the price of replacement vehicles, post-poning the replacement of high emitters.However, a combination of higher general taxa-tion on motoring and environmental standardson vehicles is always likely to reconcile restrainton car use with environmental protection betterthan discriminatory import taxation.

Transport fuel is also an internationally tradedcommodity. Having an open border and beingable to take advantage of superior fuels manu-factured in other countries makes it much easierto phase lead out of gasoline and to implementother fuel-quality improvement measures. Insome parts of the world, there is a move towardharmonizing fuel specifications to ensure mini-mal environmental standards, foster intrare-gional trade, and enhance the efficiency ofsupply. Fuel specifications in North America, theEU, and the countries of the former Soviet Unionare already harmonized, for the most part.Similar measures have been proposed in LatinAmerica.34

Public expenditure policiesThe traditional role of public sector expenditurein addressing environmental externalities cen-ters on the provision of classic public goods.These public goods include (a) maintenance ofan emissions inventory and (b) setting andenforcement of standards, including establish-ing I/M arrangements (even if these arrangementsare implemented by the private sector).

Environmental benefits are often attributed topublic transport subsidies because of modaltransfers from car to bus, or from bus to rail.Generally, these arguments should be treatedcautiously for two reasons. First, the “preferred”modes only yield benefits if they are well patron-ized. Poorly loaded buses are more environ-mentally damaging than are well-loaded cars.Second, the environmental benefit only accruesif the subsidized modes draw their patronagefrom an environmentally inferior mode, ratherthan from new trip generation. In the case ofcar–public transport transfers, the evidence isthat the cross-elasticity of demand between cardriving and public transport is low. In the case ofrail-bus transfers, which are the most common,the problems concern the financial and economiccost of securing the transfers. It is therefore wiseto draw a cautious conclusion: not that it is impos-sible to secure environmental benefit from publictransport subsidies, but that subsidy alone mightbe an inefficient means of securing that environ-mental benefit. Public transport subsidy is bestaddressed in the much wider context of improv-ing congestion and distribution, as well as in thecontext of environmental benefits. Other forms ofenvironmentally focused subsidy can representbetter value for money. As argued above, well-targeted support of premature retirement of environmentally damaging vehicles or fuel con-versions is much more likely to yield high bene-fits-per-dollar committed.

There is, of course, much more to public expen-diture decisions. Giving priority to infrastructurefor public transport and NMT may be more effec-tive in changing modal choice than may subsidies,

58 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

and may also be more consistent with the generalpoverty reduction strategy of poorer countries.

SYSTEM MANAGEMENT POLICIESSystem management policies to reduce the envi-ronmental impact of urban traffic can be dividedinto three categories. These categories are (a)those giving priority to less-polluting modes, (b)those relieving the impact by allowing traffic toperform in a more environmentally friendly way,and (c) those relieving the impact by reducingtraffic volumes.

Public transport prioritiesIn developing countries, buses are often the trans-port mode of choice for the poor, but they arefrequently highly polluting because of the manystops and starts and idle running of engines inheavy traffic. Giving priority to buses not onlyreduces their direct environmental impact butalso improves their attractiveness with respect tothe private car. It also improves their finances.The construction of separated busways, as hasbeen done in several Brazilian cities, or even ahighly integrated priority bus network, as inCuritiba, appears capable of affecting car own-ership and, more significantly, car use. It is stilldoubtful how far good bus services by themselvescan defer or deter motorization. The case for buspriorities will thus be a combination of economic,distributional, and environmental considerations.

In this context it is appropriate to consider theconcern that economic liberalization of transportoperations will have adverse environmental con-sequences. In road haulage, where liberalizationis best established, it has usually increased bothaverage vehicle size and load factor. As long asthis is accompanied by adequate enforcementof emissions standards, the effect should be ben-eficial rather than harmful. In the bus industry theopposite appears to have happened, with bothaverage vehicle size and load per vehicle falling,often in the context of poor environmentalenforcement. In the absence of a well-function-ing regulatory system, complete liberalization of

bus services, such as has occurred in Mexico Cityand Lima, has clearly increased environmentalpollution. That outcome is not inevitable, how-ever. For example, faced with this effect inSantiago, the Chilean government successfullyintroduced competition “for the market,” in theform of a competitively tendered franchisingsystem using environmental quality of the vehicleas one of the selection criteria. This highlightsthe importance of concentrating on regulatoryreform, rather than simple deregulation; this isdiscussed further in chapter 7.

Traffic managementTraffic congestion reduces average speed andincreases most emissions (except for NOx). Trafficcongestion worsens the emissions of both localand global pollutants. Increasing the averagespeed in city traffic from 10 kilometers per hour(km/h) to 20 km/h could cut CO2 emissions bynearly 40 percent. Increasing vehicle speed from12 to 15 km/h to 30 km/h in Bangkok and KualaLumpur was estimated to be equivalent toinstalling three-way catalytic converters in 50 per-cent of the cars in these cities. Experience sug-gests, however, that though it is possible toreduce congestion in the short run, this favorableresult invites greater car use in the long run, thusultimately resulting in further congestion. Studieshave shown that measures to decrease traffic con-gestion by providing more road space eventu-ally increase the volume of traffic.35

More subtle forms of traffic management may beable to reduce unit emissions rates without gen-erating extra traffic to negate the benefit.Coordination of traffic lights is generally benefi-cial. Traffic-calming devices, which slow trafficdown but do not stop it, may also result in cleaner,as well as safer, traffic. Good signage can pre-vent excessive “hunting” movements for scarceparking space.

Traffic restraintIn the industrial economies, demand for morespace in lower-density settlement, and the vehi-

TRANSPORT AND THE URBAN ENVIRONMENT 59

cle ownership and use associated with it, haveall proved income elastic, while at high incomesand low motoring costs, the price elasticity ofdemand for car travel has been low. Increasedvehicle fleet and mileage may therefore appearinevitable as economies develop. The politicalfeasibility of denying those economic character-istics has also been widely doubted.

In some of the world’s most congested and pol-luted cities, politicians are reassessing that con-servative judgment; some quite direct attackson vehicle use have been recently made for envi-ronmental reasons. Exclusion of some vehicles,selected by registration numbers from all roadson particular days, as in the “hoy no circula” (“donot drive today”) scheme, first imposed inMexico City in 1989 and subsequently copied inother cities, such as São Paulo, can result in dra-matic reductions in traffic volume during the firstmonths of implementation. In the longer term,however, the schemes may be counterproduc-tive because some households purchase an addi-tional vehicle or retain an old and pollutingvehicle that would have otherwise been replaced,in order to avoid the effect of the restrictions.The most recent variant of this strategy, the “picoy placa” (“peak-hour restrictions by license platenumber”) scheme in Bogotá, applies only topeak hours, but increases to two the number ofdays per week on which each vehicle is kept offthe road.

Further focusing restraint to exclude all vehiclesfrom particularly sensitive areas (such as pedes-trian-only city shopping centers or residentialareas) is now adopted in many countries, and isincreasingly being provided for in initial planningof new development. When first introduced itwas believed that this type of measure wouldreduce trade in the controlled areas and hence beresisted by traders. In actuality, however, itappears not to have had that effect and, whenassociated with good planning of preferentialpublic transport access, can reduce private roadtraffic over a broader area.

For the broadest impacts, economic instrumentsstill look the most promising. The World Bankenergy sector strategy argues that it should bethe objective for all countries to integrate localenvironmental and social externality costs intoenergy pricing and investment decisions.36 Theextreme case is that of Singapore, where verystrong political action to limit the stock of cars tothat deemed sustainable has been implementedthrough the auctioning of a controlled stock ofcertificates to purchase vehicles as well as by con-gestion pricing in the CBD and some of its majorfreeway accesses.37 The existence of a singlejurisdiction with the ability to use the revenuesfrom vehicle taxation to finance other support-ing elements of a comprehensive policy seemsimportant to that case. Perhaps equally signifi-cant, in the United Kingdom the legal provisionof revenues from congestion pricing to accrue tothe local authority introducing it has been thebasis for the development of new enthusiasm forthe pricing device from business as well as envi-ronmental interests.

In some industrialized countries, studies also indi-cate that, in the long run, own-price elasticity forgasoline consumption is significant enough tomake fuel taxation a potential policy instrumentfor reducing vehicle usage and kilometers trav-eled. A World Bank study concluded that a judi-cious use of gasoline tax could save the citizensof Mexico City $110 million a year more thanwould an otherwise well-designed control pro-gram with no gasoline tax.38 This might be accen-tuated by converting some of the other costs ofownership (such as insurance, parking, and vehi-cle taxation) into costs of use, although in manydeveloping countries these charges are low andoften evaded.

NMT modes are the least polluting of all modes,as well as often being the less expensive for short-distance movements. The elimination of imped-iments to NMT thus has environmental benefits,in addition to the poverty focus advantages dis-cussed in chapter 8.

60 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

CONCLUSIONS: A STRATEGY FOR URBAN TRANSPORT AND ENVIRONMENT

A number of strong conclusions arise from theabove discussion of strategic priorities.

On basic knowledgeBetter basic understanding both improves policyidentification and helps to persuade decision-makers to implement the policies:

• Local data collection and analysis on vehi-cle registrations and on levels and sourcesof ambient air pollution

• Development of better understanding of thehealth impacts of different small particulateemissions from transport (although thisdevelopment may be better left to the indus-trialized countries)

• Dissemination of basic knowledge on envi-ronmental impacts of transport modes (forexample, optimal oil/fuel mix for two-strokegasoline engines)

• Education campaigns on efficient vehicleoperation, maintenance, and so on.

On technological priorities While it is generally preferable to concentrate onperformance standards, rather than on specifictechnology preference, there are some relativelyclear-cut technological priorities:

• Elimination of lead from gasoline • Replacement of two-stroke by four-stroke

motorcycles • Elimination or cleaning up of high-mileage,

heavily polluting vehicles; the Bank can helpboth with technical assistance in these fieldsand, in some cases, with the financing ofpublic infrastructure and incentive mecha-nisms to stimulate change

• Introduction of computerized I/M regimesadministered by centralized, private sectorcontractors, subject to scrutiny to preventcorruption, and targeted initially at old andhigh-polluting vehicle categories.

On managing transport demand Technological and fiscal measures must be com-plemented by a coherent transport managementstrategy lest increasing traffic volumes swampthe beneficial effects of other measures. Thiswould include:

• Public transport investment programs,including improved conditions for walkingand cycling

• Traffic management, including rigorouslyenforced priorities for public transport in con-gested and environmentally sensitive areas

• Traffic-calming and other measures ofdemand management.

On incentive systemsBecause most critical decisions about travelbehavior are made by individuals, and are largelydriven by economic self-interest, tax levels andstructures are often decisive in determining theamount of transport undertaken, choice of mode,technology, and fuel. This emphasizes:

• Tax reform both to restrain demand fortransport to an efficient and environmentallyacceptable level and to generate efficientincentives in the choice of vehicle and fueltype, size of vehicle, and location and timingof vehicle use

• Exploration of ways to overcome limitationsof fiscal measures associated with multiplesectors and multiple objectives.

On institutionsBecause there are multiple pollutants and multi-ple types of ambient conditions, comprehensive,multi-instrument packages need to be tailored tospecific local circumstances. This requires:

• Assessment of environmental impacts as anintegral part of transport and land-use struc-ture planning

• Development of technical competence andprobity in administration as an essential pre-requisite for effective action, because of thecomplexity of these packages

TRANSPORT AND THE URBAN ENVIRONMENT 61

• Development of concerted action betweenjurisdictions and tiers of government.

The Bank can help to identify and focus on majorpolluters, and can help in the internationalexchange of experience in designing integratedurban environmental strategies. Moreover, inte-gration of transport interventions in generalmunicipal development packages may offerbetter leverage than transport-specific projects,as cities seek to reduce pollution.

NOTES

1. IPCC 1996.2. Lvovsky and others 2000. 3. Schipper and Marie-Lilliu 1999.4. Eskeland and Xie 1998.5. Eskeland and Devarajan 1996.6. For technological detail of the issues

involved, see Kojima and Lovei 2001.7. The advisory board to the U.S. National

Toxicology Program has recommended thatdiesel exhaust particles be listed as “reasonablyanticipated to be a human carcinogen.”

8. Recent studies have found a significant inde-pendent effect on premature mortality for ozoneif the important nonlinear effects of temperature(and humidity) are taken into account (Holgateand others 1999).

9. Ideally, a comprehensive urban air strategypolicy should involve the systematic collectionof air quality data and identification of majorsources contributing to the pollutants of concern.Source apportionment analysis, emissions inven-torization coupled with dispersion modeling, canbe carried out to identify which sources shouldbe targeted for control. These steps will thenenable comparison of costs and benefits of dif-ferent identified measures for improving air qual-ity, and to identify priority actions (Kojima andLovei 2001).10. Lovei 1996.11. Sulfur in diesel was lowered to 0.05 weightpercent (500 parts per million by weight, wt ppm)in 1993 in the United States and in 1996 in Europe

to meet new standards for particulate emissions.This move came after a number of vehicle tech-nology measures had substantially lowered thecarbonaceous contribution to particulate emis-sions.12. In the Uni ted States the Ai r Qual i tyImprovement Research Program of 1989–95found that high-emitting, poorly maintained vehi-cles contributed about 80 percent of total vehi-cle emissions but represented only about 20percent of the population. Improving fuel qual-ity decreased emissions somewhat but not nearlyas much as changing vehicle technology (forexample, by identifying and repairing old vehi-cles). Similarly, the European Programme onEmissions, Fuels and Engine Technologies foundthat the spread in emissions levels related to vehi-cle technologies was wider than the variationsattributable to fuels.13. In comparison with a modern, catalyzedgasoline car, a CNG-fueled car of equivalentsize has been estimated to emit approximately10 to 20 percent less CO2 and particulates pervehicle kilometer, up to 25 percent less NOx,and 80 percent less CO, nonmethane hydro-carbons, and other smog-forming emissions.For buses, compared with diesel 80S Euro 2 bus,a lean-burn CNG bus has an advantage in allthe major pollutants, including a 20 percentadvantage in global warming gases and an 85percent advantage in particulates, despite thefact that the modern clean diesel bus emits onlyone-quarter of the NOx and one-eighth the par-ticulates of the diesel bus of 1990. The stoi-chiometric CNG bus has a smaller advantageover diesel in CO and CO2 (only 10 percent) thanthe lean-burn version, but an even greateradvantage in NOx. For a fuller discussion andsources of statistics quoted here, see Gwilliam2000a.14. For example, transportation and distributiondouble the wellhead price for a refueling stationon the Atlantic coast of Colombia, but give a sta-tion cost five times the wellhead price for the cen-tral zone, where the distribution network is only20 percent utilized. The real economic cost is thuslowest in urban areas close to the gas fields.

62 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

15. For example, for the barest 12-meter low-floor bus, the extra cost of a CNG vehicle mightbe $30,000 on a price of $120,000. For the moresophisticated vehicles used in urban transport inthe U.S. and Western Europe, the basic costmight be twice that, while the incremental costof CNG would not increase proportionately. Thatmight be reduced by international unification ofregulations for homologation and safety, andeventually through scale economies in OEM. Indeveloping countries retrofitting is the norm, asthe conversion of existing vehicles may be lessexpensive, even in the long run, than prematurereplacement by dedicated new vehicles.16. LPG requires pressures ranging from 4 to 13bar, compared with 200 bar for CNG.17. At present the weight of available lead/acidbatteries is likely to account for nearly 40 percentof the vehicle mass, limiting speed to about 40km/h and range to about 55 kilometers. The costof eight batteries and additional modificationsrequired for a three-wheeler is about $1,000,which doubles the price of electric comparedwith two-stroke gasoline three-wheelers. Whilebattery technologies under development (Ni MHor Lithium Ion) may raise achievable speeds toover 50 km/h and range to over 100 kilometers,these performance levels are still much inferiorto those of existing technologies.18. ACEA (the organization of the Europeanmotor industry) and the European Commissionhave agreed that average CO2 emissions fromnew cars will be reduced by one-quarter com-pared with current levels, to 140 grams per kilo-meter by 2008. 19. For further guidance on selection of tech-nology, see TRB 1998a.20. Nitrogen oxides (NOx) are formed duringcombustion as nitrogen in the air reacts withoxygen at high temperatures. The amount of NOx

formed can be reduced by controlling the peakcombustion temperature (for example, by recir-culating exhaust gas in vehicles), by reducing theamount of oxygen available during combustion,or by converting NOx to nitrogen and oxygen-containing inorganic compounds after its for-

mation (for example, by installing three-way cat-alytic converters for gasoline engines).21. Many countries have made effective use ofdifferentiated taxation—taxing leaded gasolinemore than unleaded gasoline—to encouragethe use of unleaded gasoline and prevent theuse of leaded gasoline in cars equipped withcatalytic converters. In the absence of such afiscal policy, control programs need to be in placeto prevent permanent deactivation of catalyticconverters on a large scale as a result of misfu-eling. The effectiveness of differentiated taxa-tion depends on the extent to which fuelquality—in this case, leaded versus unleadedgasoline—is enforced.22. Unfortunately, there are many examples ofuncoordinated policies. In one country the gov-ernment is proposing mandatory installation ofcatalytic converters in heavy-duty diesel vehicleswithout taking measures to lower the sulfur levelin diesel (currently 0.7 weight percent, or 7,000parts per million by weight [wt ppm]). At such ahigh sulfur level, the life of the catalyst will beshortened, and the oxidation catalyst will merelyoxidize sulfur completely to SO3 (and sulfate par-ticulate matter will form), thereby significantlyincreasing particulate emissions. In another coun-try the government mandated catalytic convert-ers in passenger cars without specifying theemissions levels to be met—an omission thatcould defeat the purpose of the new require-ment. In other cases, catalytic converters havebeen mandated without a reliable system of pro-viding unleaded gasoline.23. Kojima, Brandon, and Shah 2000.24. For example, in June 2000 the current pricesof three-wheelers from an ex-showroom in Delhi,inclusive of all state and local taxes, are Rs. 66,500(US$1,359) for a two-stroke vehicle and Rs. 70,500(US$1,440) for the four-stroke equivalent. In July2001 no difference was discernible in the aver-age price of equivalent two-stroke and four-strokemachines.25. Typically consisting of a catalyst to oxidizeHC (hydrocarbon) and CO, and secondary air toassist conversion.

TRANSPORT AND THE URBAN ENVIRONMENT 63

26. The durability standard for catalytic con-verters in Taiwan, China, is still only 15,000 kilo-meters, although in 2000 the Society of IndianAutomobile Manufactures offered to give a war-ranty of 30,000 kilometers for all two- and three-wheeler vehicles fitted with catalytic converters.Even this is only one year’s life for a typical three-wheeler in commercial use.27. Weaver and Chan 1996.28. In a recent U.S. Agency for InternationalDevelopment (USAID)–financed program, theSociety of Indian Automotive Manufacturesundertook free emissions tests for 65,000 partic-ipants in 12 locations over a three-week period.Vehicles failing the test were given a free mechan-ical check involving carburetor adjustment, sparkplug cleaning and adjustment, and air filter clean-ing. As a result of this minor maintenance, emis-sions of HC (hydrocarbons) were reduced by 30.9percent and those of CO by 59.7 percent.29. Roth 1996.30. Empirical evidence from Spain suggests thatchanges in the I/M program may have had con-siderable effects on trends in first-time vehicleregistration (European Conference of Ministersof Transport 1999a).31. European Conference of Ministers ofTransport 1999a.32. In a pilot program implemented in 1996 inBritish Columbia, Canada, owners were offeredthe choice of cash or a one-year free transit pass(worth about 1,000 Canadian dollars) on the localpublic transport network, to replace vehicles ofmodel year 1983 or older. Fifty-two percent ofthe owners participating in the program chosethe second option.

33. World Bank 1996.34. There is a tradeoff between harmonizingfuel specifications and setting site-specific fuelstandards. Ideally, provided that the distributionsystem can handle segregation of different fuels,cleaner (and costlier to produce) transport fuelsshould be used in large cities, and fuels with less-stringent specifications should only be used inareas outside urban centers. Leakages and otherenforcement problems, as well as the logisticsof delivering fuels of different qualities to dif-ferent depots, make it difficult to implementregional differentiation cost effectively. Countriesthat import the bulk of their transport fuels typ-ically find it easier to harmonize with neighbor-ing countries than with countries that havedomestic refineries. Trade considerations(between Canada and the United States, withinCentral America, and within the European Union[EU]) are a strong driving force for harmonizingfuel specifications. In the case of developingcountries, harmonizing fuel specifications withNorth America or the EU is most unlikely to becost effective.35. The U.S. Environmental Protection Agencysuggests that up to one-half of the annual U.S.traffic growth of 2.7 percent could be a result ofconstruction of added road capacity, while areport of the British Standing Committee on TrunkRoad Assessment in 1994 concluded that increas-ing the capacity of the road network eased con-gestion only temporarily because of additionalgenerated traffic (Sactra 1994).36. World Bank 1998.37. Willoughby 2000b.38. Eskeland and Devarajan 1996.

THE SCALE OF THE ISSUE

This chapter distinguishes between the problemsof transport safety, which are defined as vulner-ability to accidental injury (usually involving atleast one vehicle as the instrument causing theinjury), and the problems of transport security,which are defined as vulnerability to intentionalcriminal or antisocial acts suffered by thoseengaged in trip making.

Recent conservative estimates suggest that, in1999, between 750,00 and 880,000 people diedas a direct result of road accidents.1 The WorldHealth Organization (WHO) puts the numbereven higher—at 1.171 million people.2 About 85percent of these deaths occurred in the devel-oping and transitional economies; about one-half were in urban areas. In addition, between25 and 35 million people were injured in roadaccidents worldwide, of which up to 75 percentwere in urban areas. The economic cost of theaccidents in the developing world has been esti-mated at $65 billion, which is approximatelyequal to the total annual aid and lending of theinternational institutions to these countries. Forthe developing countries, the economic cost of

accidents is estimated to be between 1 and 2percent of their gross domestic product (GDP).Road accidents currently rank ninth as a causeof deaths worldwide, and are expected to riseto sixth by the year 2020. Even more significant,because many of the people killed are relativelyyoung, road accidents already rank second interms of reductions in l ife expectancy. InBangladesh, it is reported that nearly 50 percentof hospital beds are occupied by road-accidentvictims.3

The security problem is less well quantified orrecognized. It particularly affects pedestriansand cyclists, but also affects people in cars andpublic transport vehicles. In extreme cases, suchas the injuries or deaths of passengers in theSouth African minibus “wars,” there is somerecord of the occurrence of events. More usu-ally the acts of personal violence or harassment—particularly sexual harassment in public transportvehicles—do not get recorded. Social surveys inLatin America have demonstrated their preva-lence (Gomez 2000).

The significance of poor safety and security istwofold. First, there is the direct injury and

URBAN TRANSPORT SAFETY AND SECURITY 65

URBAN TRANSPORT SAFETY AND SECURITY5Nearly 0.5 million people die and up to 15 million people are injured in urban

road accidents in developing countries each year, at a direct economic cost of

between 1 and 2 percent of worldwide gross domestic product. A majority of

victims are poor pedestrians and bicyclists. Fears for personal safety and security

significantly deter the use of nonmotorized transport. This burden of physical

harm that is borne by the poor can be reduced by improved road design, traffic

management, medical service, and by policy improvement. This solution requires

comprehensive action by a well-trained, committed, adequately financed, and

organizationally integrated public sector.

66 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

trauma suffered by victims. According to a recentpublic transport survey carried out in Lima, 10percent of those interviewed had been involvedin a traffic accident on public transport duringthe previous six months.4 Second, there is theeffect of the perception of vulnerability on thetravel patterns of a much wider spectrum ofpeople. The major constraint on the use of non-motorized transport (NMT) is the fear of acci-dent or attack, or of a bicycle’s being stolen whileit is parked. The diminishing trip rates now beingexperienced in some of the larger SouthAmerican cities have been partly attributed tothe sense of insecurity experienced by poten-tial passengers. While this affects all socialgroups, the most vulnerable people appear tobe those who have no alternative to an insecuremode of travel, and whose protection takes theform of withdrawal from a socially importantactivity (for example, evening education ofwomen). When a wage earner in a poor family isbadly injured, the whole family economy maycollapse, because there is usually neither insur-ance compensation nor a social security safetynet to protect them (box 5.1).

SAFETY

The World Bank’s concern with transport safetyis not new. Periodic reviews during the past twodecades have shown an increasing proportion ofprojects with an explicitly stated safety objectiveand safety components, with nearly one-half oftransport projects having safety elements in thelatest cohort examined.5 However, most of theseefforts arose as part of the design of infrastruc-ture improvements or traffic management sys-tems, with safety audits becoming a commonpart of new transport infrastructure projects.Improvement of the accident databases has beenincluded in many projects. The total cost of theseactivities was estimated as only about 1 percentof total project costs.

Relatively few projects have had transport safetyas their primary objective. The Mexico HighwayRehabilitation and Traffic Safety Project includedinstitutional strengthening, research, and train-ing, and a $14 million expenditure on black-spotimprovements. The Buenos Aires Urban TransportProject included an even larger sum to eliminate

BOX 5.1 ROAD ACCIDENTS AND THE POOR IN SRI LANKA

A study of pedestrian road accidents in Sri Lanka showed that only 5 percent of accidents reportedto the police resulted in any form of compensation. Even where compensation was paid, the aver-age amount was 500 rupees (about US$5).

A thousand pedestrians and cyclists are killed in the country each year. The maximum compen-sation payable is only 100,000 rupees (about US$1,040), compared with the unlimited sum payablefor damage to property, which may run to millions of rupees. The victims’ loss of earnings is oftennot covered by motor insurance. Moreover, the legal procedure in making such a claim is toolengthy for the poor to even attempt.

The Motor Traffic Act should clearly stipulate a reasonable and fair compensation for pedestri-ans and cyclists. There is also need for provision of free (or affordable) legal aid to assist the poorand uneducated in making such claims, and for counseling and rehabilitation in the case of theseverely injured.

Source: Kumarage 1998.

URBAN TRANSPORT SAFETY AND SECURITY 67

at-grade road crossings of the urban and subur-ban rail network, to speed up rail travel, and toreduce accidents. The most comprehensive effort,however, was the collaborative effort with thePHARE program of the European Union (EU) toimprove traffic safety in six countries of Easternand Central Europe from 1992 to 1994.

The outcomes of these interventions, and othersof the same type, have been variable. A reviewof 25 projects in the early 1990s found them tobe equally divided between the categories ofsuccessful, partly successful, and unsuccessfulinterventions.6 Physical measures were usuallyimplemented, but their effects seldom measured;enforcement efforts were often initially success-ful, but rarely sustained; and road-user educa-tion and legislative reform efforts were mostsuccessful when carried out by a well-establishedcoordinating body, such as a national safety coun-cil. Above all, few projects were judged to haveimproved institutional capacity to undertake traf-fic safety activities beyond the project itself. Ofgrowing concern, also, has been the protectionof vulnerable road users, with measures to protectNMT being an important part of recent urbantransport projects, in China in particular. The coreweaknesses have been the poor level of aware-ness and commitment of governments and anunwillingness to create a strong enough institu-tional focus for a concerted, sustained effort.

Despite the horrifying statistical evidence, itremains difficult to persuade governments to givepriority to road safety either as a transport prob-lem or as a health problem. For example, a studyof road safety in Ethiopia estimated that theannual cost of road accidents was the equivalentof 40 million Ethiopian birr ($4.8 million) per year.However, a suggestion that 2 percent of a newlyestablished road fund (equivalent to 400,000 birr,or $47,450) be spent on a comprehensive crash-reduction program each year was rejected by thegovernment.7 This reflects a certain sense of fatal-ism and apathy about the problem, nurtured bythe belief that, because of the strong human ele-ment in accident causation, there can be no well-

established, cost-effective intervention packagesfor accident prevention, as there are to addressother scourges of morbidity and mortality.

To redress that apathy, the magnitude and natureof road accidents must be properly understood.Governments must be convinced that effectiveaction is possible and that institutional arrange-ments can be put in place so that necessaryactions can be effectively implemented. For thatreason the Global Road Safety Partnership, estab-lished as the result of a World Bank initiative, hasconcentrated its early efforts on mobilizing theprivate sector and civil society to assume theirresponsibilities in road safety, increasing aware-ness of the nature of the problem, and identify-ing a limited number of pilot initiatives which canshow that something can really be done aboutit. This section concentrates on the main ele-ments of understanding the phenomenon, policyformulation, infrastructure design, traffic man-agement, medical policies, and institutions.

UNDERSTANDING THE PHENOMENONA primary source of policy neglect has been theabsence of reliable evidence on the magnitudeand nature of the problem of transport safety.Road accident fatalities and serious injuries havelong been known to be substantially underre-ported in official police statistics in developingcountries, and should be adjusted up by 10 per-cent, at the very least.8 The situation is worse withrespect to injury-only accidents.9 Even within theinjury-only accidents recorded in hospital statis-tics, there is likely to be some bias. For example,females represented only one out of every sevencasualties in urban Zimbabwe, probably due tothe reduced ability of low-wage earners (oftenincluding women) to afford hospital treatment.

The impact of road accidents is concentrated onsome c lasses o f vu lnerab le road users .Pedestrians account for more than twice the pro-portion of those injured in developing, as com-pared with industrialized, countries. Drivers andpassengers of motorcycles and three-wheelmotor vehicles account for fewer than 10 percent

68 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

of those injured in developing countries, but upto two-thirds of those injured in some East Asiancities, such as Kuala Lumpur. Public transport pas-sengers, particularly those traveling in the back ofa commercial truck or pickup truck, are very vul-nerable. In many countries drivers of trucks andbuses have particularly bad accident records.

While males between the ages of 16 and 54account for the majority of injury from accidentsin all countries, about 15 percent of those killedin developing countries are children, which is amuch higher proportion than that in industrial-ized countries. Although the police rarely collectincome data, recent interviews of pedestriansinvolved in road accidents in a number of coun-tries show that the poor were disproportionatelyaffected. The damage is even higher where theinjured are the main income earners in a house-hold. Studies in Bangladesh and Zimbabwe haveshown that 80 percent of those injured have familymembers who are entirely dependent upon them.

The location of accidents also varies significantlybetween countries. The majority of urban acci-dents in industrialized countries occur at inter-sect ions, whi le most urban accidents indeveloping countries are reported to occurbetween intersections. Relatively few accidentsoccur where there are any traffic controls, includ-ing traffic police. This is partly because in theabsence of effective development control, unre-stricted access to main roads increases the riskof a collision. It is also partly attributable to the dif-ferent mix of vehicle types that are using theroads; this is particularly seen in the juxtaposi-tion of motorized and NMT users, who are morevulnerable between intersections, where speeddifferences are greatest.

The absence of adequate accident statistics isimportant, not only because it diverts attentionfrom the seriousness of the problem but alsobecause it hinders the search for, and selection of,appropriate remedies. Identifying the most vul-nerable locations, types of accident, and types ofperson involved is the basis for road-safety policy

design. Introduction of an effective system of acci-dent recording and analysis is thus a very high pri-ority for international assistance. Because accidentanalysis would be useful to a range of agencies(including the police, the judiciary, insurance com-panies, car manufacturers, and traffic manage-ment agencies), the case can be made for theanalysis to be done as independently as possible,perhaps by a road research institute. Alternatively,it should be a function of the traffic managementagency. Computer programs developed for analy-sis in Denmark, the United Kingdom, and else-where are already in use in many countries. Butthat analysis is only as good as the attention torecording of accident detail in the police records.A critical part of the development of an accidentanalysis capability is to persuade police chiefs tocollect, process, and transfer to the responsibleagency the data needed for traffic safety analy-sis—rather than only those needed for legal pur-poses—and to train their staff accordingly.

POLICY FORMULATIONIn most industrialized countries, increases inroad accidents have been associated with theincrease in car ownership and usage. A largeproportion of deaths and injuries occur to vehi-cle occupants. Many countries, but most notablyAustralia, Japan, and the United Kingdom, havedeveloped comprehensive programs to reducethe incidence and severity of road accidents;these programs are based on a combination ofengineering, enforcement, and education.Urban traffic safety is also a current priority ofthe EU.10 Measures include improvements ininfrastructure design (which are often informedby black-spot analysis), vehicle characteristics(particularly compulsory installation and use ofsafety belts [seatbelts]), and driving behavior(such as blanket speed limits in urban areas andcampaigns to discourage “drinking and driv-ing”). High-level goals (such as Sweden’s VisionZero goal of no deaths from road crashes) aretypically adopted and advertised at the nationallevel, but are made operational at local levelsby municipal and other highway authorities.These programs have been supported by a high

URBAN TRANSPORT SAFETY AND SECURITY 69

level of agreement and coordination betweendifferent authorities under different ministriesand with different budgets.

Similar comprehensive road-safety programs canbe successful in developing countries. For exam-ple, in the early 1990s the Asian DevelopmentBank assisted the government of Fiji to developa broad national road-safety action plan thatreduced road deaths by 20 percent. However,because of the differences in traffic compositionand consequently in the characteristics of acci-dents and of the most vulnerable groups, differ-ent policy instruments are likely to be of highpriority in developing countries.

INFRASTRUCTURE DESIGNThere is no doubt that good design of roadi n f r a s t r u c t u re c a n h e l p s u b s t a n t i a l l y.Improvements in road surface and horizontaland vertical alignment at black spots has provedvery effective in a number of cases.11 Clear def-initions of, and implementation of, a road hier-archy can help to match the use and operatingspeed of roads to their immediate environment.Much is already well known about measures toprotect the pedestrians and cyclists, who arethe most vulnerable road users. Proper provi-sion of footways, controlled signals for at-gradepedestrian crossings, grade-separated cross-ings, pedestrian-only areas, and segregatedbicycle lanes and tracks are all effective and, incomparison with most infrastructure, relativelyinexpensive.12 For example, on the basis of pilotprojects in East Africa, undertaken as part of theSub-Saharan Africa Transport Program, it wasargued that serious urban pedestrian and bicy-cle traffic crashes can be reduced significantlyby a suitable program of road (and intersection)redesign and traffic calming.13

That type of experience needs to be generalizedand disseminated. Considerable effort hasalready been put into the preparation of designmanuals for safe road infrastructure. In 1991 theBritish Overseas Development Agency (now theDepartment for International Development)

funded the preparation of a manual entitled“Towards Safer Roads in Developing Countries,”which has been widely disseminated, both in theEnglish and Spanish versions. The AsianDevelopment Bank has funded road-safety guide-lines for the Asia and Pacific region, and coun-try-specific road-safety engineering manuals havebeen developed in a number of countries, includ-ing Bangladesh, Indonesia, Kenya, and Malaysia.The Inter-American Development Bank hasundertaken similar work in Latin America.14

Incorporation of a safety audit of road (and traf-fic management system) designs by an inde-pendent consultant is likely to be a cost-effectiveway of avoiding the need for black-spot improve-ment programs later.

Safety is a necessity rather than a luxury, but con-ventional methods of cost-benefit analysis maymake it look like a luxury unless the benefits ofimproved safety are appropriately valued. Thereis, of course, an understandable reluctance toattribute a money value to the saving of life or tothe reduction of pain and grief. Certainly inter-national comparisons of the value of life are invid-ious. For allocations of committed funds inexplicitly safety-oriented projects, the issue canbe evaded by use of cost-effectiveness analysisto compare alternative project designs. But whereit is a matter of safety-related design componentsof investments, for which the bulk of impacts aretime- or vehicle operating–cost savings, the omis-sion of safety valuation will make safe designappear as an uneconomic luxury. It is thereforesuggested that all governments insist that safetybenefits be attributed a value that appears rea-sonable in terms of local conditions. More-detailed advice on how they might approachevaluation is available.15

Road-safety plans and action programs havebeen prepared in many countries, usually byexternal consultants piggybacked onto otherprojects. While these programs have been verybroad, they were often led by road-safety pro-fessionals, with only limited support from the localenforcement and legal authorities.

70 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

TRAFFIC MANAGEMENTWhere there is no independent traffic-safetyanalysis unit, the safety functions of a “traffic man-agement agency” generally commence withretrieval of accident data from the traffic police.There is often no systematic, periodic transfer ofdata from traffic police to the traffic managementagency, with data retrieved on an ad hoc basisto resolve particular accident problems. Amethodical approach requires that the trafficmanagement agency obtain data on a regularbasis and that procedures be established withinthe traffic management agency to allow the acci-dent data to be analyzed to determine prob-lematic sites, periods, groups, trends, and so on.Various proprietary accident-analysis softwareprograms are available, but any simple databasesoftware package can be used.

Although a traffic management agency may havea separate road-safety group with the responsi-bility to analyze accident data, promote safetyprograms, and review schemes, safety shouldbe regarded as an integral part of any traffic man-agement scheme design and should be animportant evaluation criterion governing theacceptance of any scheme or measure. In somecountries, such as the United Kingdom, all butthe simplest of schemes are subject to an“Independent Safety Audit.” This involvesscrutiny by traffic management designers whowere not involved in the original scheme plan-ning and design. In some developing cities, it isacknowledged that there may be few experi-enced traffic management staff and there maybe a lack of resources for hiring consultants.Nevertheless, the savings in social costs fromthe introduction of “safe” schemes should morethan offset costs; the independent safety audit isworth consideration as part of the normal designprocess.

It is generally accepted that in industrialized coun-tries, the three most common causes of fatalitiesand injuries are (a) excess driving speed, (b) driv-ing under the influence of alcohol, and (c) inad-equate protection of vulnerable persons in

accidents. At the national level there should beenforced systematic policies for dealing witheach, while at the local level those policies shouldbe rigorously enforced.

Speed limits and controls are powerful instru-ments to reduce the severity of accidents. Onlocal roads in European cities, a wide range ofphysical traffic-calming measures for speed con-trol has been used effectively. Typical measuresinclude:

• Pedestrian refuges that narrow the effectiveroad width

• The control of vehicle overtaking (passing)and prevention of vehicles from reachinghigh speeds

• Road humps, to reduce vehicle speed• Road narrowing, to prevent heavy vehicles

from using a road or to restrict movementof vehicles to one direction at a time

• Chicanes, to force vehicles to follow a tor-tuous route and thus reduce speed

• Raised intersections, comprising a plateauor flat-topped road hump built across anentire intersection

• Plantings, to change the perceived width ofa road in order to encourage vehicles toreduce speed.

On main roads, speed limits must be enforcedby the traffic police by various means—directmeasurement by radar guns, static or mobilecamera enforcement, following vehicles, and soon. Traffic calming can also reduce traffic speeds,especially if carefully related to the hierarchy ofroads. On main roads, effective devices includepositive signs and road markings emphasizingspeed limits, rumble devices, bar markings, roadtexture and color on the approaches to criticallocations (intersections, pedestrian crossings, andso on), and adjustment of intersection traffic-signal timings to control and maintain a desiredsafe speed of traffic progression. However, someof the more extreme physical traffic-calmingmeasures used on local roads might add to acci-dent hazards if introduced on main roads.

URBAN TRANSPORT SAFETY AND SECURITY 71

The strict enforcement of stringent national stan-dards on drinking and driving is the basis forreducing the second serious cause of accidents.The right to perform random tests assists enforce-ment, but may be a platform for corruption insome countries. Holding employers of profes-sional drivers, as well as the drivers themselves,responsible is also a powerful inducement toeffective control, especially in public transportcompanies. Above all, it is important that it is theoutcome (reduction of drunken driving) and notany particular procedure (for example, daily med-ical inspection of drivers as routinely required inmany countries of the former Soviet Union) thatis subject to control.

In industrialized countries, efforts to protect per-sons in accidents have concentrated on seat beltand airbag installation and use. In some middle-income developing countries, the emphasis hasbeen on the use of crash helmets by bicyclistsand motorcyclists. In many poorer countries,however, the real issue is the protection of pedes-trians from motorized vehicles; the provision ofadequate sidewalks, barriers, and road-crossingfacilities is most important. While the provisionof pedestrian bridges or tunnels may offer thegreatest potential protection, it may not be themost effective measure, especially where thecrossings involve arduous detours or aredesigned as a potential operating ground forthieves.

MEDICAL POLICIESThere is considerable evidence that the lack ofadequate medical facilities contributes to thehigh level of fatalities in developing-countrycities.16 Many lives could be saved if medicalattention were provided within the hour imme-diately following an accident (the “golden hour”).This requires the improvement of emergencyservice response time, which can often beimproved at modest cost by the following:

• Strategic positioning of emergency servicecenters (perhaps first aid stations at fuel sta-tions)

• Provision of an emergency telephonenumber

• Establishment of a control center • Use of ITS (Intelligent Transport Systems)

applications for efficient service control• Setting up an emergency medical services

committee• Provision of first aid training• Creation of a mechanism, possibly funded

by insurance companies, to cover costs ofminor expenses in bringing injured personsto the hospital

• Upgrading hospital emergency rooms anddepartments.

INSTITUTIONSIn many countries too many different agenciesand institutions have some responsibility for roadsafety for it to be viewed as the primary respon-sibility—and hence institutional priority—for anyof them. Moreover, some of those institutions,notably the police departments, often have sucha bad image in developing countries that bothcitizens and international institutions may beloathe to support them.17 It is therefore impor-tant that emphasis should be placed on the devel-opment of an institutional focus for transport(particularly road) safety. Emphasis thus needs tobe given to developing an institutional responsi-bility for coordination of safety efforts at a veryhigh level, while at the same time enhancing thecommitment of the interested line agencies(police, traffic management, health, and educa-tion), at both the national and local levels. Directresponsibility of the national road-safety agencyto the prime minister’s office is a device used tofocus attention on, and obtain satisfactory com-mitment to, road safety in countries such as Indiaand Vietnam. Parallel institutional arrangementsat the municipal level, with direct responsibilityto the head of the city government, have beensuccessful in prosecuting urban road-safety cam-paigns (box 5.2).

Funding arrangements for road safety need par-ticular attention. Most cities finance safety meas-ures out of limited departmental construction

72 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

and management budgets. In Vietnam somededicated road-safety funding is obtained at alocal level from traffic fines. Of the new genera-tion of road funds that have been developed inrecent years in Africa and elsewhere, only thosefunds in Ethiopia are known to specify safetymeasures, along with road-maintenance activi-ties, as the responsibility of the fund.18

Other sources of funding are clearly required.One source of increasing interest in a number ofindustrialized countries is through contributionsfrom insurance company premiums. However,since a large proportion of road vehicles are oper-ated without insurance in many developing coun-tries, this is only likely to offer a viable source offunds in the wealthier and better-governed coun-tries. In a very few cases private sector supporthas been mobilized. In Delhi, Indian vehicle man-ufacturers Maruti Udyog Ltd. have sponsoredinterceptor patrol vehicles. These patrol vehicleshave played a prominent role in traffic law

enforcement and resulted in a considerableincrease in revenue from fines—however, it is notpossible for the police to reinvest the revenuefor road-safety activities. This sponsorship cameabout following the brokering of a partnershipbetween a nongovernmental organization, thetraffic police, the Delhi state government, andMaruti Udyog (Aeron-Thomas and others 2002).

SECURITY

Personal security while engaged in transport activ-ity is an increasing problem throughout the world.In a sense, this is not a transport problem but asymptom of a much wider social malaise. But theinescapable need to undertake travel to pursueessential activities of life—such as work, educa-tion, health care, and so on—may force peopleinto situations where they are most vulnerable toattack, with only a limited ability to adjust activ-ities to avoid or ease their vulnerability (box 5.3).

BOX 5.2 SAVING LIVES IN BRASÍLIA

As a planned new city, Brasília has an extensive road network, which in 1995 sustained an averagetraffic speed of 40 kilometers an hour (km/h), twice the national urban average, but which alsoexperienced 11 deaths per 1,000 vehicles. On the recommendations of a joint working group ofthe secretariats of public safety and transport, in July 1995 the governor established by decree atraffic safety program, entitled “Peace within Traffic.” The aims of the program included:

• Control of excess speeding• Control of driving under the influence of alcohol• Tighter traffic rules enforcement• Improved medical assistance to accident victims• Improved road infrastructure safety features• Vehicle safety inspection and control• Pedestrian, cyclist, and public transport priority.

Several secretariats were involved in implementing this high-level activity, which was supportedby an energetic press campaign, as well as by intensive efforts to involve civil society. Between1995 and 1997 the number of deaths per 1,000 vehicles fell from 11 to 6.6, and the emphasis andbenefit has been subsequently maintained.

Source: Affonso, Rezende, and Vitor 1998.

URBAN TRANSPORT SAFETY AND SECURITY 73

Threats to security of person and property may beclassified into four main types.

a. Theft by stealth, which is largely a functionof crowding on public transport vehicles, butwhich may also involve the unattended park-ing of bicycles and other vehicles.

b. Theft by force, which can occur in crowdedplaces but is more likely to occur in situa-tions where the victim is relatively isolated.Theft by force includes vandalism and vio-lent physical attack.

c. Sexual harassment, which with differentdegrees of violence can occur in eithercrowded or isolated situations.

d. Political and social violence, which may havesome transport significance (such as the attackson South African commuters traveling by rail,bus, or minibus) or for which the transport vehi-cle may simply be an opportune location.

In each case, while the origin of the problem maynot lie primarily in transport conditions, ques-tions arise about the planning and managementof transport facilities and services.

Theft by stealth is the most common manifesta-tion of this problem, and is the most difficult to act

against, but usually, fortunately, is the least trau-matic of the phenomena. Passengers in vehiclescan be frequently reminded of the need for cau-tion, and of the best ways to secure themselvesagainst theft. Automatic prosecution and exem-plary sentencing of those caught can also be adeterrent. Provision for secure parking of bicy-cles has been an important element of policiesthat support bicycle ownership, in some coun-tries. Electronic surveillance may be effective instations but less so (and more expensive) oncrowded vehicles.

Theft by force, because it is more likely to occurin less-crowded locations, is more susceptible toelectronic surveillance, which, however, is onlylikely to be effective if accompanied by adequatearrest-and-arraignment arrangements. The exis-tence of a specialist transport policing force hashelped in rail and metro systems in industrializedcountries, but is less likely to be affordable forthe fragmented bus sector.

Vandalism, which is a form of property theft, andunruly behavior toward passengers are commonin poorly managed public transport operationsin both developed and developing countries.Management changes or institutional reform canrapidly reduce vandalism. For example, graffitialmost disappeared from the New York subwayonce public transport management took deter-

BOX 5.3 CRIME, VIOLENCE, AND DIMINISHING MOBILITY

In a number of major cities in Latin America, such as São Paulo, the number of trips undertakenper day have been declining in recent years, and it has been suggested that this is at least partlya consequence of declining security, particularly in the evening hours when trip rates have declinedmost. This interpretation is supported by evidence from a survey of poor households in Ecuador.In a six-month period in 1992, one in five women in Cisne Dos was robbed on a bus, and one intwo women had witnessed such an attack. There was a drop in the use of public transport at night,and an increase in the relatively safer small trucks operated by the informal sector. For those whocould not afford the alternative, travel was curtailed. The lack of safe transport during off-peakhours has caused girls, generally from the poorest families, to drop out of night schools.

Sources: Henry 2000; and Moser 1996.

74 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

mined action. In Buenos Aires, service on thegovernment-run suburban railways had becomevery irregular and unreliable in the 1980s andearly 1990s, and windows were broken, seatsslashed, and passengers harassed by rovingbands of vandals. When railway services were pri-vatized in 1994, the first action of the new con-cessionaires was to introduce controllers(supported by government security guards) oneach train—in part to control fare evasion and inpart to establish a safer environment. Within fouryears the number of passengers had doubled,to a large degree because of the improved safetyassociated with train travel.

Perhaps even more than the public transport pas-senger, the pedestrian is increasingly likely tosuffer violent attack. This may occur after dark aspart of a robbery or, in the case of women, sexualassault. It can occur in business or residential dis-tricts, but is most common in low-income settle-ments controlled by gangs in the absence of aviable police presence. Again, the poor suffermost, because they are vulnerable to physicalattack when walking from bus stops to home.Travel by taxi is expensive—and often not evenan option when drivers refuse to drive into dan-gerous neighborhoods. For example, in Caracas(República Bolivariana de Venezuela) it wasreported that people missing the last safe oppor-tunity to return home in the evening are obligedto stay overnight at their places of work. In somecountries, such as Ghana and South Africa, theftby violence from cars or of cars when stationaryor slow moving has been a problem, againstwhich drivers tend to protect themselves by theequally dangerous procedure of ignoring trafficsignals, particularly after dark. Civilian neighbor-hood patrols, common in some industrializedcountries, may also play a role in improving safetyfrom violent physical attack in the developingworld. Police bicycle patrols, which have becomepopular in the United States, have also shownpromise in Venezuela.

Sexual harassment can be reduced by the pro-vision of women-only vehicles in situations where

the density of movement makes this feasiblewithout loss of service availability. Examples ofthis include buses in Bangladesh, India, and SriLanka; and coaches on some trains in the Mexicometros. In Karachi, Pakistan, women-only com-partments in buses are physically separated fromthe larger conductor-controlled compartment ofthe vehicle. Given increasing personal awarenessof the problem, a commercial response is begin-ning to emerge in some Latin American minibusand taxi markets that specifically protect vulner-able travelers. This response might be stimulatedat very little cost by government encouragementand some externally supported experiments.Harassment by male transport staff unwilling tomake allowance for the difficulties of women inentering or leaving moving vehicles might beaddressed by use of mixed-gender crews onpublic transport vehicles.19 Female police offi-cers have become effective, and have a reputa-tion for being tough and incorruptible, in theenforcement of traffic rules in La Paz, Bolivia, andLima. Women also might have a wider role inresponding to issues of sexual harassment.

Political and social violence often finds a focusin burning buses or destroying traffic signals, evenwhere there is no transport-related stimulus. Thereare also some transport-specific origins of vio-lence. Bus and rail passengers in South Africawere allegedly targeted in order to coerce themto ride the black-operated minibuses. Minibuspassengers were also frequently caught up in mur-derous struggles between competing operators.These types of insecurity are particularly suscep-tible to actions designed to regularize and givelegally defensible property rights to operators offranchised services. Economically motivated policyreform in urban transport operations may thushave a very significant security payoff. The bene-fit accrued depends on the regulations beingenforceable, and on being actually enforced, bylegitimate authorities, and not by mafias. Forexample, in Medellin, Colombia, gangs control-ling a low-income area extorted protection moneyfrom bus operators serving the area; funds wererecovered from the fares of the poor passengers.

URBAN TRANSPORT SAFETY AND SECURITY 75

Some general points may be made in conclusion.Increasing criminality in many developing cities isa symptom of a much wider social malaise. Whileit affects the transport behavior of everybody, itis primarily the poor who suffer when essentialtrips for work or education are curtailed. Lack ofsecurity also frustrates environmentally motivatedattempts to reduce the need for car travel whenchildren can no longer safely walk or take the busto school, and many people are obliged to goby car or taxi when even a short walk may havebecome too dangerous. To some degree, secu-rity in public transport might be improved byestablishing minimum regulations on service qual-ity. There are some technical fixes to improve per-sonal security for pedestrians, such as betterstreet lighting and use of video or CCTV (closedcircuit television) monitoring of public spaces,20

but ultimately this is a function of much broaderand more complex issues, such as social cohe-siveness and the tradeoff between police powerand human rights concerns.

CONCLUSIONS: A STRATEGY FORURBAN TRANSPORT SAFETY ANDSECURITY

The development of a strategy for urban trans-port safety should include:

• Development of national road-accident sta-tistics data collection and analysis capabil-ity

• Incorporation of safety elements in all trans-port infrastructure projects by the incorpo-ration of a mandatory safety audit in thedesign process

• Incorporation of estimation and evaluationsafety benefits of improved designs in allinfrastructure projects, using values deter-mined by government in collaboration withlocal safety agencies

• Development and associated training of stafffor specific road-safety coordinating agen-cies or councils, both at the national and themunicipal levels

• Specification, clear signing, and enforce-ment of maximum speed limits for differentroad categories in urban areas

• National-level specifications, advertising,and enforcement of limits for blood-alcohollevels for vehicle drivers

• Financing of specific safety-related infra-structure investment (such as the financing ofinfrastructure for NMT, or the railway cross-ing investments in Buenos Aires) based onthe identification of vulnerable groups andlocations

• Involvement of police in road safety, such asthe collaboration between police and traf-fic management departments in black-spotanalysis in Seoul

• Involvement of medical authorities in jointplanning for improved accessibility to med-ical facilities for victims of accident trauma

• Inclusion of compensation provisions andliabilities in motor traffic and associatedinsurance legislation

• Creation of high-level committees withresponsibility for road safety in all major cityadministrations

• Development of plans for financing safetyactivities as part of transport strategy plansin all major municipalities.

With respect to security, serious effort remainsnecessary both to analyze the nature and signif-icance of insecurity in the urban transport sector,and to devise policy instruments to counter it.That might include:

• Collection and analysis of data on personalsecurity in the transport sector

• Development of an awareness of the prob-lem, together with the commitment of policeauthorities to arrest, and the courts to appro-priately penalize, delinquents

• Development of franchise conditions givingincentives for improved attention to secu-rity by public transport operators

• Including street lighting—designed to improvepedestrian security—in street improvement,and particularly in slum-upgrading projects

76 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

• Strengthening public participation in proj-ects—particularly those dealing withimprovements at the neighborhood level.

NOTES

1. Jacobs and Aeron-Thomas 2000.2. WHO 1999.3. TRL and Silcock 2000.4. Gomez 2000.5. Amundsen 1996.6. Ross 1993.7. TRL and Silcock 2000.8. A TRL study in Colombo, Sri Lanka, in 1984

showed that fewer than one-quarter of hospital-reported road traffic accident casualties wererecorded in police statistics. The problemremains. In Karachi, 1999 police statistics showedonly 56 percent of the fatalities and only 4 per-cent of the serious injuries attributed to road acci-dents in hospital statistics. Even hospital recordscan be a poor source for accident research. InBuenos Aires most deaths are attributed solelyto a medical condition (broken skull, for instance)and not to the cause of that condition.

9. The number of urban injuries reported forevery fatality in official statistics in 1999 was 160 inGreat Britain, 22 in Zimbabwe, but only 3 in Dhaka.10. The DUMAS (Deve lop ing UrbanManagement and Safety) project involves col-laboration of research teams in nine countries toproduce a framework for the design and evalua-

tion of urban safety initiatives (EuropeanCommission 2000).11. For example, the Amman Transport andMunicipal Development Project, funded by theWorld Bank, planned improvements at 15 inter-sections. Crash reductions of 98 percent wereachieved between 1984 and 1990 at the first twointersections treated. However, these were theonly two implemented, and only 18 percent ofthe planned expenditure on traffic managementwas achieved.12. Cracknell 2000.13. Koster and de Langen 1998.14. Gold 1999.15. TRL 1995. 16. TRL and Silcock 2000.17. For example, there is widespread concernabout the provision of equipment—such as cars,cameras, and other enforcement devices—whichmight also be used for less legitimate, non–traffic-related purposes.18. Even in this case, the only safety measurefinance so far has been the upgrading of trafficsigns in Addis Ababa. Since 91 percent of trafficaccidents in Addis Ababa involve pedestrians,the provision of new traffic signs is unlikely to bea significant safety benefit.19. Gomez 2000.20. Some cameras used for traffic control havealso been used for other purposes, but this raisesbroader questions of personal privacy and pos-sible misuse of government power.

THE URBAN ROAD SYSTEM 77

ELEMENTS OF ROAD STRATEGY

Most urban transport, be it public or private, pas-senger or freight, motorized or nonmotorized, inrich countries or in poor, uses the road system.Roads also provide rights-of-way for utilities, com-munications infrastructure, and the everyday inter-action of the population. The competitiveness ofcities, on which the wealth and welfare of theirpoor as well as their rich increasingly depends ina global economy, requires efficient road trans-portation.

Despite its economic significance, the road systemis usually managed in a fragmented and uneco-nomic way. Decisions on the management, main-tenance, and expansion of road infrastructure restwith separate public sector agencies, while thoseconcerning operations on that infrastructure fallpredominantly to the private sector. Increasedprivate demand for infrastructure is not reflectedin the price charged for it and does not yield anyincrease of revenue or fiscal commitment to sat-isfy it. The juxtaposition of extremely buoyant pri-vate demand (demonstrated by high levels ofcongestion) and a cash-starved infrastructure-supply agency is commonplace. Inadequate roadmaintenance is pandemic. The most important

requirement of a strategy for roads in develop-ing countries is to link those private and publicsector decisions in an economically rational way.

A strategy for roads in urban areas to make thislinkage must address a number of major issues,including:

• Road maintenance organization and finance• Traffic management to improve traffic system

capacity, quality, or safety, or a combinationof all three

• Demand management to secure maximumsocial value from network use

• Infrastructure expansion planning andappraisal.

While these aspects are primarily concerned withthe growth-creating effects of roads, the directpoverty impacts should also be addressed.

ROAD MAINTENANCE

Cities usually have a massive investment in theirroad systems, which are often very poorly main-tained. World Bank–funded projects typically showvery high returns on maintenance expenditures,

THE URBAN ROAD SYSTEM6Most urban transport is based on the use of roads. Congested road infrastructure

damages the city economy, increases environmental pollution, and harms the

poor by slowing road-based public transport. Particularly in larger cities,

however, it may be neither socially nor economically acceptable to balance supply

and demand solely by increasing road capacity. A strategy for roads must

therefore concentrate on the movement of people, rather than the movement of

vehicles, through traffic and demand management, as well as on the provision

and maintenance of road infrastructure.

78 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

yet there is a persistent tendency to underfundmaintenance. In the interurban highway context,this has been confronted by the development ofuser-managed “second-generation road funds”financed through fuel duty surcharges and otherdirect charges for use.1 Such funds do not exist inall countries, however, and even where they do,

they do not necessarily solve the problems ofurban road maintenance (box 6.1).

The creation of a systematic and secure basis forthe maintenance of economically and sociallyviable roads should be a high priority of urbantransport strategy. Where, as is commonly the

BOX 6.1 FINANCING URBAN ROAD MAINTENANCE IN THE KYRGYZ REPUBLIC

The Kyrgyz Republic—a small nation in central Asia with a population of approximately 5 mil-lion—is facing a serious urban road maintenance problem. Bishkek, the capital, with an estimatedpopulation of 1 million, has a road network of 730 kilometers, of which nearly 90 percent is paved.Lack of routine maintenance and deteriorating drainage conditions have resulted in street sur-faces that are uneven, heavily cracked, and extensively potholed. Much of the road network needsrehabilitation or complete reconstruction. Other, smaller Kyrgyz cities have similar problems.

A road department of the city corporation is wholly responsible for road maintenance within thecity. Work can, in principle, be carried out either by force account or by private sector contrac-tors. In practice, private sector capability remains limited. Asphalt and concrete plants are stilloperated by the city. Moreover, the embryonic local private sector road construction industry isplagued by the lack of a stable level of contracts and the reluctance of the cities to contract outservices.

Providing adequate financing of road maintenance and rehabilitation is critical. The recently estab-lished Kyrgyz Republic National Road Fund is permitted by law to allocate up to 10 percent of itsresources to urban areas, but has so far allocated much less than this because the needs of thenational road network are considered more pressing. Cities are in no better position to fund roadmaintenance from their own sources. Transfers from central government make up 60 to 75 percentof city budgets, but are mostly earmarked to pay for school and health-care worker salaries. Lessthan 2 percent of total city expenditures go to roads—about $0.12 per capita per year. As it hassuch limited authority to raise revenues, Bishkek has resorted to ad hoc lotteries, bazaars, andfestivals to finance road maintenance.

In connection with a $22 million International Development Agency credit to help cities catch upon the backlog of maintenance and road rehabilitation of arterial roads carrying the bulk of motorvehicle and public transport traffic, the Kyrgyz government has set up a working group of city-and central-government officials to find ways to make funding of road repairs more sustainable.Agreement has already been reached that National Road Fund revenues could be increased, witha higher percentage coming from user fees (such as gas and vehicle taxes). The ultimate chal-lenge will be to empower cities to generate and retain adequate revenues to fulfill their statutoryresponsibility for road maintenance and rehabilitation. Urban road finance in the Kyrgyz Republicis thus essentially part of wider national road finance and municipal finance issues.

Source: R. Podolske, from World Bank project files.

THE URBAN ROAD SYSTEM 79

case, maintenance is financed from generalmunicipal revenues, some general allocation stan-dard should be established as a benchmark. Incountries with national road funds, there shouldbe adequate criteria and process to ensure thatan appropriate part of the fund is devoted tourban roads. That has not been the case in someof the early road funds.

In terms of funding sources, the second-generationroad fund philosophy is that road infrastructureshould be maintained with funds collected fromusers. Ideally, direct or indirect charges to usersshould be sufficient to cover the full costs of traf-fic, including environmental impact and roadmaintenance. Distance- and weight-related licens-ing fees might be the most accurate instrumentsfor allocating maintenance costs between cate-gories of road user, although allocations of fuelduties or congestion charges might be proxiesfor those fees.

In practice, many countries do not have such roadfunds, and road maintenance responsibilities areallocated between levels of government accord-ing to whether the traffics using specific roads areconsidered national, regional, or local. Where sucharrangements are in place, it is essential that thedelineation of responsibility between authoritiesis clear, and that the funding arrangements aresteady and sustainable, with substantial discretionfor municipalities to raise funds locally, if necessary.

Implementation also needs improvement in manycountries. A maintenance culture must be estab-lished and nurtured in the municipal institutions.Maintenance management systems should beemployed to plan and budget for required main-tenance on a systematic basis related to surveysof road condition, distinguishing routine, peri-odic maintenance, and rehabilitation or recon-struction.2 This establishes the foundations forthe development of a private maintenance con-tracting capability and even possibly for thedevelopment of urban road maintenance con-cessions, as have recently been concluded inMontevideo (Uruguay) and Bogotá.

TRAFFIC MANAGEMENT

The objective of urban traffic management is tomake the safest and most productive use of exist-ing (road-based) transport system resources. Itseeks to adjust, adapt, manage, and improve theexisting transport system to meet specified objec-tives. By maximizing the efficiency of existingfacilities and systems, capital expenditure can beavoided or deferred to gain time in which todevelop longer-term policy measures. At thesame time, traffic safety can be improved andthe adverse impact of road traffic on the city envi-ronment reduced.

Many traffic management instruments, such asefficient traffic signal systems, increase the effi-ciency of movement with negligible adverse sideeffects. But traffic management policy alsoinvolves choices. Priority given to pedestriansmay reduce capacity for vehicles or may adverselyaffect bus operations. An integrated traffic man-agement scheme will require compromisesbetween the competing interests of various usersof the road and traffic systems. It is hence not justa technical issue, but must be clearly directed inharmony with the overall urban transport strat-egy. In practice, in most cities both traffic engi-neers and traffic police tend to concentrate onkeeping the traffic moving. As a result, roads arewidened, motorcars (automobiles) receive prior-ity, and pedestrians and bicycles get crowdedout. In that process the street system getsrearranged to benefit the generally richer carusers at the cost of the generally poorer bus users,bicyclists, and walkers. If it is a policy objectiveto focus on the needs of the poor, this objectiveshould also motivate traffic and demand man-agement.

TRAFFIC CONTROLThere are numerous traffic management tools,including on-street parking management andcontrol, traffic circulation design, traffic signalsystems, public transport (bus) priority, and theenforcement of traffic regulations. These toolscan be applied not only to speed the movement

80 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

of cars but also, if desired, to give priority topedestrians, bicycles, and other nonmotorized(NMT) or commercial vehicles.3

The combination of a coherent circulation systemdesign and an efficient traffic signal system iscentral. Good signal control technologies arereadily available. Moreover, the rapid pace ofdevelopment of intelligent traffic system tech-nologies offers relatively poor countries thechance to leapfrog to the latest technology, justas some have in telecommunications. Such sys-tems are increasingly robust and need little main-tenance, but they do require planning, someminimum maintenance, and a willingness to usethem. Unfortunately, many large cities lack eventhe basic skills to develop and maintain thesesystems, with the result that, as in Bangkok, traf-fic control reverts to what the police can achieveat a single intersection.

Traffic management can be applied at a specificsite; for example, improvements may be neces-sary at a key intersection to ensure consistenttraffic capacity along a route, or improvementsmay be made at an intersection to resolve acapacity or severe accident problem. However,traffic management is most effective if appliedover an area (for example, a corridor, local area,or town center) in order to develop a consistenttraffic management regime. Hence measures arebest combined in comprehensive packages toensure that traffic problems are not simply trans-ferred to new conflict points, that there is syn-ergy between the various interventions, and thatusers are consistently presented with the same“messages,” thereby improving the likelihoodof observance of traffic regulations.

PUBLIC TRANSPORT PRIORITYIn most cities in the developing world, buses arethe backbone of the motorized transport systemand will remain so for the foreseeable future. Inpoorer cities the bus share of the market may beexceeded by paratransit (publicly available pas-senger transport service that is outside the tra-ditional public transport regulatory system).

Regular travel by taxi or motorized rickshaw-taxi(as in some cities in India) tends to be by thebetter-off; the car is not a mode available to mostpoor people. Even in cities where suburban railis predominant, such as in Mumbai, or wheremetros exist, most mechanized journeys are usu-ally still by public road passenger transport.

At peak loading, a bus may carry 30 (or more)times as many passengers as a car, in only threetimes the road space. The primary objective oftraffic management should therefore be toimprove travel for people and not necessarily forvehicles. Not only will this lead to the most effi-cient use of scarce road space but the policy willhave a positive impact on poverty. Such anapproach will tend to emphasize:

• Measures to assist public transport gener-ally and bus priority measures in particular

• Pedestrian facilities• Bicycle facilities.

In many Latin American cities, traffic manage-ment already seeks to improve on-street busoperations through busways or bus lanes.Extensive bus lanes and busways exist in manycities, with the most notable being Curitiba, PortoAlegre (Brazil), São Paulo, Bogotá, Lima, Quito,Santiago, Mexico City, and León (Mexico). ManyLatin American cities also have pedestrianizedareas or streets, particularly in the city centers.Road paving has been undertaken in some poorareas to provide bus access, and most roads areprovided with pedestrian footways. Elsewhere,the situation is less encouraging. In Asia fewdeveloping cities—including, as far as is known,only Bangkok, Manila, Madras, and KualaLumpur—have allocated extensive road spaceto bus priority and busways. In the latter threecities, enforcement of with-flow bus lanes (asopposed to the contra-flow bus lanes that con-stitute most of the Bangkok lanes) remains prob-lematic.

The lack of bus priority outside Latin America ispartly associated with the absence of a conven-

THE URBAN ROAD SYSTEM 81

tional bus system. The prevalence of paratransitmakes bus priority difficult to implement due tothe high volume of relatively small vehicles, suchas the jeepneys in Manila. In some cities, such asHanoi, Vietnam, bus services are embryonic; evenwhere bus services exist, they may be regardedby users as the transport of last resort, as in somecities in China. In other cases, such as Moscow,bus transport is regarded as a subsidiary modeto the metro and hence given little preferentialtreatment on trunk routes. For these reasons thereis a lack of understanding of the objectives of buspriority, and politicians are unwilling to committo measures that adversely affect the (rich) pri-vate car users. Because of its low status, there isa lack of trained and experienced professionalstaff with sufficient vision to appreciate the ben-efits of road-space reallocation to buses. Theproblems of enforcement of “complex” bus pri-ority schemes that require vehicle selectivity (thatis, buses versus others) are viewed as too great toovercome.

The trend in industrialized cities is the reverse ofthat in developing cities (outside Latin America).Road-space reallocation to buses from cars on ahighly preferential basis is increasingly acceptedin many richer industrialized cities, particularly inWestern Europe. As in Latin America, acceptancefor bus priority has been brought about by betterplanning, better information to decisionmakers,better enforcement or legislation, creation of abetter image for buses, and better disseminationof information on the benefits of bus priority.Busways in such cities as Curitiba and Quito arehighly valued by the population (most of whomhave low incomes), and mayors spearheadingbusway schemes have subsequently won electionto even higher offices.

IMPLEMENTATIONIn many cities the high rates of traffic growthmay quickly catch up with the initial congestionrelief of traffic management measures. Trafficmanagement must not be seen, therefore, as apanacea for urban transport congestion, butrather as a component of a broader strategy

also involving public transport and demandmanagement. Traffic management must not beseen as a “one-shot” intervention, either—rather, it should be seen as a continuous process,adapted and adjusted to meet the changingtraffic situation. Hence, the emphasis should beon creating a favorable institutional environ-ment for the effective operation and adaptationof traffic management measures, and on fos-tering the technical skills to implement them,rather than merely on the financing of specificschemes.

The main traffic management activities to be per-formed at a city or city-region level are shown intable 6.1. The major institutional problems relat-ing to road provision and management concernthe functional separation of responsibility foractivities between municipal and district levelsin hierarchical systems, and the coordinationbetween local traffic management and overalltransport strategy.

Achieving a balance between competing trans-port modes and interests relies on competenttransport institutions working within a clearlydefined framework of responsibilities. While thereis no single ideal or model institutional frame-work for traffic and transport administration, acity must have an organizational framework thatdeals with the basic functions of strategic trans-port planning; infrastructure planning, design,and construction; road maintenance; public trans-port planning and procurement; and traffic anddemand management, including enforcement.Alternative ways of organizing these functionsare discussed in chapter 11.

Some traffic management functions can be con-tracted out to the private sector. The contract-ing may be conventional (such as for themaintenance of traffic signals) or include abroader range of functions. Consultants maybe contracted to conduct a large part of thetraffic management process—for example, todevelop and implement a comprehensive cor-ridor plan—although the executive must ulti-

82 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

mately remain responsible and accountable forthose functions. While this may reduce the needfor implementation staff at the city level, it doesnot alleviate the need for a trained core of staffto determine traffic management policy and tomanage the consultants. Issues arise over theextent of the consultants’ or contractors’

responsibility for letting (that is, awarding)implementation contracts and thus responsi-bi l i ty for public funds. However, such anarrangement may assist cities in developingcountries, since it may be easier for a city toobtain finance for short-term consultants thanfor the staff levels required in-house.

TABLE 6.1 FUNCTIONS AND RESPONSIBILITIES OF A TYPICAL TRAFFIC

MANAGEMENT AGENCY

Area Functions and responsibilitiesTraffic management Formulate and implement citywide traffic management policies to comply with objectives policy defined by the city council to include, at least, such areas as determination of (a) a functional

road hierarchy, (b) the appropriate balance between transport system users (private transport–public transport—NMT vehicles–pedestrians), (c) priority programs for action, and (d) “five-year” investment plans.

Traffic research Survey, monitor, and evaluate all traffic and accident data to enable trends to be identified, problems quantified, and traffic management plans and improvements prepared.

Traffic management Plan, design, implement, monitor, evaluate, fine-tune, and continuously update traffic schemesplans and improvements and policies to realize the agreed-on traffic management policy. The program would cover all

motorized road-based modes (cars, public transport, trucks, and so on) and all NMT modes (pedestrians and cycles). Plans and improvements would range from simple intersection improvements, or marking and signing programs, through to far-reaching citywide strategies such as extensive bus priority or pricing. Accident programs and countermeasures would be included.

Traffic control devices Plan, design, install, operate, and maintain all traffic control devices including (a) traffic signal systems, including computer-controlled systems; (b) road markings; (c) road signs; and (d) enforcement devices (cameras, and so on).

Traffic regulations Formulate traffic regulations to realize the proposed traffic management plans and improvements, for enactment by city government and for enforcement by the traffic police.

Parking management Prepare off- and on-street parking policies and programs, including approval for the location of, and access to, parking areas proposed by others. Parking enforcement and administration, where paid parking applies, would be carried out by a separate “parking authority,” or the equivalent.

Approvals and Evaluate and advise city government on all schemes (such as new roads) and developmentscoordination (developed both by public and private sector agencies, and including major new land or

building developments) that have a significant traffic impact, to ensure that they are consistent with agreed-on traffic policy.

Consultation Consultation with the public and stakeholders on traffic policy and on the impacts of specific schemes and measures.

Budget Preparation of an annual budget for submission to city government for (a) implementation of traffic management plans and improvement schemes, (b) traffic operations and maintenance of control devices, and (c) the continuous work of the traffic management agency itself.

Notes: Not all functions would be carried out by the traffic management agency itself. For example, functions such asmaintenance of traffic control devices and signals would most commonly be contracted out. In this case, the agency wouldassume the functional responsibility of supervision. The potential for contracting out significant elements of the functions toconsultants is discussed below.Source: Authors.

THE URBAN ROAD SYSTEM 83

DEMAND MANAGEMENT

The economic rationale for demand manage-ment is that if the price directly incurred by trav-elers in making journeys is lower than the fullcost of the journey, then some trips will imposea net cost on the community. The full costs of ajourney include both the personal costs incurredby the traveler (vehicle-running costs, fuel, park-ing, and so on), and the social costs imposed bythe traveler on the community through the jour-ney’s contribution to congestion, the increase inaccident potential, and the polluting effects onthe environment. As the costs imposed by a trav-eler on others vary by location, time, and trafficconditions, so, ideally, should the chargesincurred by vehicle users also vary. The objec-tive of demand management should be to securethe total level of traffic, and its distributionbetween modes, locations, and times of day thatwould exist if traffic by each travel mode wereto be charged prices equal to its full marginalsocial cost.

In order to achieve that objective, all demand-management tools aim to increase the costs oftravel, either explicitly through charges (parking orcongestion charges or fuel prices) or implicitly bylimiting movement through traffic restraint meas-ures. Demand management and restraint in traf-fic volumes may be realized by a range of pricingmeasures, many of them touching on nationalpolicy (including, for instance, fuel pricing).

Parking controls for demand managementParking control and pricing are the most commonlyapplied demand management measures in bothindustrialized and developing-country cities. At itsmost basic, parking policy in many developingcities is limited to the control of on-street parking(usually simple parking prohibitions on main roads)to avoid obstruction to moving traffic. But it canalso have wider restraint potential.

While parking controls may have some effecton vehicle ownership, the usual aim of restraint

policy is to reduce car use by regulating park-ing space and by positive allocation of avail-able parking space among different groups ofusers—usually seeking to deter car commutingfor work. The effectiveness of parking as arestraint measure in developing countries iscompromised by the ability of the rich to keeptheir vehicles on the street in the charge of theirhired drivers, and by the fact that significantparts of the parking stock in many cities are notin the control of the traffic authority. While, inprinciple, publicly available but privately pro-vided car parking can be required to charge“restraint-level” parking rates with the excessprofit taxed away from them, in practice thisenforcement measure is very difficult to imple-ment in developing-country cities.

Even more difficult problems arise in respect of“private nonresidential” parking in the owner-ship of private sector companies or governmentagencies. Withdrawing the right to use theseparking areas not only creates a political prob-lem but also legal problems in some countries,where the private nonresidential parking mayhave been constructed in response to city park-ing standards for construction permits. There arestill many developing-country cities where, evenin city centers, minimum rather than maximumparking standards are set.

Despite these weaknesses a comprehensive park-ing policy is likely to be the starting point fordemand management in most cities. Parking feesare the least contentious of user charges, andmost cities have some form of parking policy. Thenumber, location, and price of on-street parkingspaces can be controlled. Publicly available off-street parking and private nonresidential park-ing capacity expansion can be limited, andcharges regulated to prevent subsidized park-ing. Policies for these categories of parking con-sistent with general policy objectives should beincluded in all transport strategy plans. The inte-gration of parking standards within developmentcontrol strategies, such as in the Dutch ABCpolicy, is a good example of this.4

TRAFFIC RESTRAINTTraffic demand can be restrained through the useof a range of physical measures such as:

• Limiting entry to environmentally sensitivecity centers to “essential” traffic only, as inTehran, Islamic Republic of Iran

• City-center pedestrianization, as in BuenosAires and Budapest

• Cell systems that limit traffic within the citycenter by creating a system of cells betweenwhich there is no direct access for cars, as inGothenburg (Sweden) and Bremen (Germany)

• Road-space reallocation to high-occupancyvehicles

• Traffic-calming measures (see chapter 5).

While the use of such measures may reduce traf-fic in the target areas (usually a city center), theyare unlikely to reduce demand overall unless cou-pled with other measures, such as parking policyand encouragement to public transport.

Among the most popular restraint measures areschemes that limit use of vehicles on specific daysaccording to their registration plate number. Thesehave been introduced in many cities—includingAthens (Greece), Bogotá, Lagos, Manila, MexicoCity, Santiago, São Paulo, and Seoul—for bothcongestion and environmental reasons. There areobvious risks to the “odds and evens” policy andits variants. They may encourage shift to taxi useor an increase in the number of vehicles owned,and induce more trips by permitted vehicles thanwould otherwise have been made. Nevertheless,they have worked in the short term (Bogotáreports a 20 percent increase in average travelspeeds). Above all, they have achieved accept-ance by the public as a demonstration of com-mitment by government to reduce congestionand related air pollution, and have proved lessdifficult to enforce than one might have expected.Particularly if well designed to discourage peakuse and if coupled with public transport improve-ments, as in Bogotá, they can at the very leastgive the government some breathing space todevelop even more effective policies.

PRICINGPhysical restraint measures have hitherto provedmore acceptable than have direct charges for roaduse, both in industrialized and in developing coun-tries. Even in industrialized countries, however,their effectiveness appears to have beenexhausted; and some countries, such as theUnited Kingdom and the Netherlands, are nowplanning the introduction of direct charges.Singapore—now a rich country—is moving towardcharges for vehicle use rather than charges andother controls on their ownership. In the few casesin countries of the Organisation for Economic Co-operation and Development (OECD) where directcordon or area congestion prices are charged,part or all of the revenues have been earmarkedfor public transport improvements. For cities indeveloping countries, which lack resources tofinance urban transport, the introduction of directcharges might thus be expected to have a doubleattractiveness as a source of finance as well as aninstrument of restraint. The incorporation of directcharges within an integrated urban transport pric-ing policy is such a potentially important con-tributor to sustainable urban transport that it isdiscussed in detail in chapter 10.

One aspect of restraint is particularly important.Both theoretical research and practical experi-ence indicates that combinations of car restraintand public transport improvement appear to workbetter than either in isolation, at least in theireffect on travel to city centers. A coherent policyis therefore likely to include a combination ofmore than one measure.

INFRASTRUCTURE PROVISION

There is no magic number defining the appro-priate level of transport infrastructure provision,which depends on the type of transport and land-use system that is being designed, as well as thetopography and economy o f the c i t y.Predominantly auto-dependent cities in theUnited States may devote as much as 35 percentof their urban space to transport infrastructure.

84 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

THE URBAN ROAD SYSTEM 85

European cities, which now have high car own-ership levels, allocated 20 to 25 percent to roadsbefore motorization. These cities now attemptto maintain system performance through trafficand demand management, and afford a degreeof physical and fiscal priority to public transportsystems. In contrast, Asian cities, many of whichdevote only 10 to 12 percent of their urban spaceto roads, do so by accident rather than by choiceand do not have the appropriate policies in placeto maintain efficient transport in the city with theroad space available.

It is clear that cities with only 10 to 12 percent oftheir area devoted to movement cannot supportunfettered motorization. Moreover, once the cityfabric is established, it becomes increasinglyexpensive, and both socially and environmen-tally disturbing, to superimpose substantial addi-tional road infrastructure. Furthermore, wherecongestion is already suppressing demand,increasing capacity may generate such a largeamount of extra traffic that congestion is reducedmuch less than anticipated. Cities that havealready grown to a size and density at which sub-stantial traffic congestion exists may thus havelost forever the option of road-building their wayout of congestion.

It is also the case, however, that as cities expand,so must their road systems. Cities need a basicamount of circulation space adequate for theirsize in order to operate efficiently (see discussionin chapter 2). Early planning of, and reservationof space for, transport infrastructure is thus astrategic requirement. Essential elements in thesuccessful integration of land-use and transportfacilities in well-planned cities (such as Curitibaand Singapore) have been the definition and pro-tection of the basic transport infrastructure cor-ridors. The requirement is not simply for roadsfor motor vehicles—pedestrians and NMT alsoneed space for circulation. Some of the mostattractive urban environments are those in whichthe circulation of motorized and NMT traffic hasbeen separated and clear priorities in the allo-cation of space determined.

The strategy on road provision thus calls for acareful balance of judgment. On the one hand,the dynamics of city expansion must be recog-nized, because even the most environmentallyaware cities only achieve their ends by adequateallocation and good planning of movementspace. On the other hand, heavily congestedcities cannot road-build their way out of con-gestion, and attempts to do so will almost certainly fail or carry very heavy social and envi-ronmental penalties. Only a strategy that recog-nizes these constraints and combines necessaryroad-capacity adjustment to support city expan-sion with clear and strict policies on the alloca-tion and use of road space is likely to be able toreconcile city expansion with livability.

HIERARCHY AND SHAPEAppropriate structuring of the road network isimportant (table 6.2). In many cases the city lead-ers want ring roads (China) and other primaryroads and expressways (Moscow) when majorproblems arise from the absence of an appro-priate structure and quantity of local distributioncapacity (Bangkok, Manila, and Jakarta). It is truethat limited-access arterial roads can handle sev-eral times more vehicles per hour in a givenamount of space than can mixed-use roads. Butit is also necessary to provide for pedestrian cir-culation and local distribution of traffic, as wellas for longer distance trunk movements. Thesefunctions do not mix well, and a given amountof road space will always give better perform-ance if it is classified and managed hierarchicallyto separate functions.5 This type of road classifi-cation is often a basis for the allocation of finan-cial responsibility for provision and an importantaid to effective distribution of maintenanceresources.

The shape of networks is also important. The clas-sic “ring and radial” design of many Western citiesevolved over time, with the ring roads beingadded as city centers became congested and theneed to provide alternative routes for intracitycrosstown movement became apparent. Theinnermost ring around the central area could be

86 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

TAB

LE 6

.2

TY

PIC

AL

FU

NC

TIO

NA

L C

LAS

SIF

ICA

TIO

N O

F R

OA

D A

ND

PA

TH

NE

TW

OR

KS

Segr

e-Pe

rcen

tage

Se

para

te

gate

d of

D

irect

Ty

pica

l M

otor

Pr

inci

pal

NM

T bu

s st

reet

la

nd

desig

n ve

hicl

e C

lass

ifica

tion

func

tions

faci

litie

sfa

cilit

ies

kilo

met

ers

acce

sssp

eed

park

ing

Com

men

ts

Pede

stria

n pa

ths

Safe

ped

estr

ian

circ

ulat

ion

Yes

No

n.a.

No

rest

rictio

nsn.

a.Pr

ohib

ited

Esse

ntia

l acc

ess

to o

ff-ro

ad p

rope

rty

Cyc

le p

aths

Sa

fe c

ycle

circ

ulat

ion

Yes

No

n.a.

No

rest

rictio

nsn.

a.Pr

ohib

ited

Pref

erab

ly c

ontin

uous

sys

tem

Loca

l str

eets

Land

and

pro

pert

y Si

dew

alks

N

ot

60–8

0N

o re

stric

tion

30–4

0 km

/hPe

rmitt

edTh

roug

h tr

affic

sho

uld

be d

isco

urag

edac

cess

desi

rabl

e re

quire

d

Col

lect

or s

tree

tsLi

nks

loca

l and

art

eria

l Si

dew

alks

N

ot

5–10

Gen

eral

ly

40–5

0 km

/hLi

mite

dTh

roug

h tr

affic

dis

cour

aged

stre

ets

usua

lly

requ

ired

unre

stric

ted

nece

ssar

y

Art

eria

l str

eets

Inte

rcom

mun

ity a

nd

Side

wal

ks

Des

irabl

e 15

–30

Acc

ess

only

to

50–7

5 km

/hLi

mite

d or

Ty

pica

lly th

e ba

ckbo

ne o

f urb

an s

tree

t in

trac

ity m

ovem

ent

man

dato

ry;

whe

re b

oth

maj

or tr

affic

pr

ohib

ited

syst

embi

cycl

e bu

s an

d ge

nera

tors

tr

ack

if ge

nera

l de

man

d tr

affic

w

arra

nts

leve

ls h

igh

Free

way

s an

d Lo

nger

dis

tanc

e N

one

Des

irabl

e n.

a.N

o la

nd a

cces

s 75

km

/h

Proh

ibite

dG

rade

-sep

arat

ed in

ters

ectio

nsex

pres

sway

sex

tra-

and

if

bus

perm

itted

or m

ore

intr

amet

ropo

litan

vo

lum

es

traf

fichi

gh a

nd

rout

e su

bjec

t to

cong

estio

n

n.a.

No

t ap

plic

able

.N

ote

: km

/h =

kilo

met

ers

per

ho

ur.

Sour

ce: A

utho

rs.

THE URBAN ROAD SYSTEM 87

designed to discourage crosstown traffic frompassing through the center and to allow moreroad space for local traffic, public transport, NMTvehicles, and pedestrians inside the ring. Someform of inner ring (coupled with appropriatefreight access arrangements) is usually an inte-gral part of major pedestrianization schemes. Thecritical element in such concepts is the manage-ment of traffic within the ring to prevent the capac-ity released from being filled by further newlygenerated through traffic. This was the basis onwhich the Bank was involved in financing of ringroads in Shanghai and Guangzhou. The main dif-ficulty in this approach is to ensure that traffic ordemand management measures are actually intro-duced, and then maintained over time.

Outer-ring roads, on the periphery of a city, maykeep crosstown traffic out of the city where routesfor crosstown traffic are inadequate, or they formpart of the regional or national highway network.The critical element is not their form (ring) but theireconomic and environmental function (by-pass),and they should be judged on these grounds. Evenwhere they are justified functionally, they may, asin Warsaw (Poland) and Budapest (Hungary), fuelauto dependency by attracting major shoppingdevelopments to the periphery, where access ismuch better by private than public transport. Theyshould thus also be appraised in terms of the con-tribution that they make to the broader land-useand transport development strategy for the city.

Other structural considerations may arise. Gridnetworks, common in U.S. cities, spread cross-town traffic across parallel and alternative routes,but cannot avoid congestion in the downtown(central area) locations with the greatest con-centrations of commercial activity. To avoid this itis likely to be less expensive and less disruptiveto strengthen the hierarchy of the grid, ratherthan to superimpose new ring roads.

APPRAISING CAPACITY EXTENSIONSIt is observed above that an adequate basic infra-structure is essential to the efficient working ofthe city but that it may not be possible to eliminate

congestion in established large cities by road-building. This observation still begs the questionof where and when capacity should be added.

The classic test concerns the elimination of “bot-tlenecks,” where the lack of capacity at one pointin a system causes congestion and delay at theapproaches to that point. Sometimes great ben-efits can be obtained inexpensively throughenhancing the continuity of a network by remov-ing a bottleneck or filling a missing link. Primafacie bottlenecks may appear to be a systemdesign deficiency and an obvious case wherecapacity expansion is necessary. Of course,though, bottlenecks are not design faults in bot-tles. There they exist intentionally in order torestrict excessive flow at the point where it is mostacceptable (in the bottle), and not at the pointwhere the flow cannot be accommodated (in theglass). The same logic may be applied to roadsystems. Increasing flow into a city center thatcannot accommodate further traffic may makethings worse rather than better. Only if the totalsystem performance is improved is the elimi-nat ion of bott lenecks a sens ible pol icy.“Opportunistic” measures to eliminate bottle-necks need to be very carefully appraised toensure that the measures are not simply shiftingtraffic congestion to another point in the system.

The conventional way to address these issues isto attempt to compare costs and benefits in arational, quantitative, framework. While this tech-nique was originally developed for appraisinginterurban investments, the necessary technicaladjustments for the appraisal of capacity expan-sion in congested cities have been part of thestandard literature for 30 years. New trafficinduced by a system improvement does implysome private benefit to those whose trips werepreviously suppressed, but it also reduces theextent to which existing traffic will benefit fromextra capacity, and it generates an additionalenvironmental impact. The net effect of allowingfor generated traffic may thus be to reduce theestimated benefit of capacity expansion. A gooddemand analysis and evaluation model may be

capable of considering this potential benefit, butwhere such sophisticated techniques are notavailable, the most obvious simplification—toassume a fixed matrix of trips—will probably sig-nificantly overvalue capacity expansion.

Another theoretically well-analyzed phenome-non is the impact of traffic and demand man-agement on the economic value of capacityexpansion. The possibilities of improving the effi-ciency of existing infrastructure by traffic man-agement, and of restraining traffic by demandmanagement, should theoretically be consideredin specifying the base case against which toassess additional capacity. In practice, however,it is too easy to presume that better manage-ment of the transport system is impossible, andhence too easy to evaluate proposed capacityexpansions on a basis that consistently overesti-mates the benefits of improvements.

Other considerations are of more recent concern,and are less well established. Effects on NMT usersof the road infrastructure are typically ignored inconventional road investment project appraisals.At the technical level, the poverty focus alsorequires that benefits to the poor should be givena higher value in economic appraisals than is oftenthe case.6 It is desirable that, at the very least, theorder of magnitude of such issues be assessed.

The cost-benefit methodology is still used at atechnical level and therefore must be amendedto avoid the most obvious biases. The valuationissues for environmental externalities, theabsence of a solid understanding of the feed-back effect of road building on land-use changesand urban form, and the distributional effects ofinvestments mean that the calculated economicrate of return is inadequate as a sole technicalcriterion. Economic evaluation therefore needsto be embedded in a multicriteria framework,designed for a participatory and politically frankdecisionmaking process.

One of the strongest but most diffused protec-tions against overinvestment is the involvement

of the concerned public in consultation or partic-ipatory design arrangements. While there is adanger of a nihilistic “not-in-my-backyard” syn-drome developing, capable of blocking good aswell as bad developments, procedures of consul-tation and of public inquiry into major schemesforce planners and engineers to think much morecarefully than might otherwise occur. It followsfrom this that the data collection and modelingshould also evolve to suit the participatory process.

PRIVATE FINANCINGPrivate finance has been mobilized for urban tollroad construction in major cities such as Bangkok,Buenos Aires, Kuala Lumpur, and Manila. Insofaras such roads are tolled at levels that service thefull resource costs of the construction, they wouldappear to be consistent with the policies advo-cated hitherto. Insofar as they also succeed inreducing congestion on the untolled public roadnetwork, they might be viewed as unambiguouslybeneficial.

That conclusion may not hold, however, for anumber of reasons. If urban toll roads generateextra traffic on an inadequate secondary roadnetwork (as in Bangkok and Kuala Lumpur), thenew toll roads may actually increase congestionand, thus, reduce welfare. The environmentalimpacts, both of infrastructure and extra traffic,are typically uncharged except insofar as they areinternalized in the design requirements for newroutes. There may also be contingent costs onthe public purse for improving the feeder net-work that are unaccounted for (as in Manila). Theconclusion is thus not that private finance is unde-sirable but that, because of the system effects,private sector schemes should be subject to thesame overall planning, economic, environmen-tal, and social analysis as they would be if under-taken in the public sector.

POVERTY FOCUS

The poverty impact of urban roads, as that of allother elements of urban transport, can be divided

88 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

THE URBAN ROAD SYSTEM 89

into the indirect effects, arising through theireffect on economic growth, and the direct effects,arising through their effect on the livelihoods ofpoor individuals.

THE GROWTH EFFECTThe economic viability of the industrial and trad-ing base of cities is important to the poor whofind their employment there. Whereas in indus-trialized countries, peak-period congestion ispredominantly caused by private cars, the sameis not true everywhere. Many major cities indeveloping countries are the terminal nodes ofnational and international transport networks.Some are historic port cities, the dominance ofwhich has been enhanced during the period ofcolonial rule. In these cities freight traffic is oftena signif icant cause of urban congestion.Concentration of traffic to and from major urbancommodity markets, as in Dhaka, may have thesame effect as the concentration of port flows, asin Manila. Having a strategy for the location ofmajor freight traffic generators, and their asso-ciated movements, is therefore an important partof an urban transport strategy in developing-country cities. Urban road strategy must also belinked to, and reflect, changes in the nationaleconomy, as is the case in some of the transi-tional economies, where dramatic changes inindustrial structure and trade links have shiftedfreight from rail to road and necessitatedchanges in urban transport structure.

DIRECT POVERTY IMPACTSThe poverty reduction potential of road invest-ments can be enhanced by the selection of typesof road project for inclusion in programs con-centrating on the road infrastructure essential tothe basic needs of the poor. These needs can beidentified and focused on in several ways.

The first way of focusing policy on the poor isto look at where they live. In many countries thevery poor live in informal housing developmentsor slums that have no paved or even preparedroad access. Because they lack road access, theyare not served by formal public transport, and

are in fact often not served by any mechanizedtransport at all. They are forced to walk long dis-tances, often in insecure surroundings. TheWorld Bank has frequently financed paving ofroads to improve access to poor areas, partic-ularly in Latin America and Africa (box 6.2). Inthe context of increasing security, lighting maybecome an important aspect of basic safe-access provision.

A second way of focusing on the poor is to look atthe modes of transport that they use. For the verypoorest, this usually means walking or other NMT.Too often the allocation of street space for walk-ing and NMT vehicles is inadequate to allow safemovement. Moreover, as motorization increases,even this limited street space tends to deteriorate.For example, while during the 1970s and 1980s theprovision of segregated cycling capacity was astandard design feature in new urban arterial roadsin Chinese cities, such attention to cyclists nowappears to be declining. Segregated infrastructurefor cycling and pedestrian access is being financedin cities as diverse as Lima, Accra, Manila, Bogotá,Shanghai, and Liaoning (China). For other groupsof the poor, and in other countries, the poor areheavily dependent on road-based public trans-port. Making sure that road space is available forroad-based public transport can be facilitated intwo ways. First, where roads have deteriorated,priority for rehabilitation may be put on the partsof the primary road network used for public trans-port (such as in a recent project in the KyrgyzRepublic). Second, where road space is scarce,priority can be given to public transport vehicles,not only by bus lanes and busways but also in moresophisticated ways, such as the incorporation ofpriority for public transport vehicles in signal con-trol systems.

The third focus might be on how the poor maketheir living. In very poor rural areas, it has beencommon practice to consider labor-intensive roadconstruction and maintenance processes, whichtend to be easier to manage and which producea relatively higher quality product where the roadstandards required are low. The same emphasis

90 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

has not been normal in urban areas, althoughexperience in South Africa does indicate thatthere may be equivalent possibilities in urbanareas.

Finally, a focus may be on the incidental impactsof road development on the poor. There is astrong temptation to route new road infrastruc-ture through areas where land is inexpensive,which is often land occupied by the poor. That islikely to be the case, in particular, where the poorare illegally occupying publicly owned land. Thestringent requirements of the Bank for resettle-ment planning and compensation stem from aconcern for this kind of adverse spin-off effect.

CONCLUSIONS: A STRATEGY FORROADS

Roads are a necessary component of urban infra-structure. Inadequacy of current road capacity tocarry current traffic results in congestion, damageto the city economy, increased environmentalimpact, and, often, particularly harsh impact on thepoor. In larger established cities, however, it maynot be socially or economically acceptable to bal-ance supply and demand solely by increasing roadcapacity. A strategy for roads must thereforeencompass traffic management and demand man-agement, as well as the provision and maintenanceof road infrastructure. A strategy must thus include:

BOX 6.2 PAVING ROADS IN LOW-INCOME AREAS IN BRAZILIAN CITIES

Unpaved roads in low-income areas in Brazilian cities, many of which used to be impassable inwet weather, were serious impediments to the access of buses and emergency vehicles. In São Pauloalone there were 800 kilometers of unpaved bus route in the early 1980s. To remedy this an exten-sive program of paving in low-income areas was included in the First Brazil Urban Transport Project.Under this project there was a tendency for the design agencies to produce excessively elaboratedesigns for drainage and basic pavement structure. Cost and time overruns resulted.

Because of these overruns, the Empresa Metropolitana de Transportes Urbanos in São Pauloundertook a comprehensive study of low-cost paving; the result was a low-cost paving manualand an economic feasibility study. In parallel, the national body Empresa Brasileira dos TransportesUrbanos (EBTU) commissioned a study of regional experiences in low-cost paving, which con-cluded that the benefits from an extensive bus-route-paving program would be substantial, andestablished guidelines for the selection of roads in a paving program.

The Third Brazil Urban Transport Project included a $63 million component—the (Programa dePavimentacao de Baixo Custo em Areas de Baixa Renda (PROPAV) program—to pave 500 kilometersof bus route in low-income areas between 1981 and 1984. EBTU would be responsible for select-ing the roads and supervising the execution, and would embody its experience in a revision of themanual. Local labor and local materials were extensively used. The program was considered sosuccessful that it was extended to over 1,000 kilometers, not all of which were Bank financed.

Similar problems existed in other Latin American and Caribbean countries, and programs werelaunched in Chile, Jamaica, Peru, and Mexico. In some cases, where impassability due to rain wasnot a problem, even simpler surfacing was used to permit extended coverage. The most recentdevelopments include local community participation in the selection of segments for inclusionin a project in Lima.

Source: World Bank project files.

THE URBAN ROAD SYSTEM 91

For road planning• Initial planning and reservation of space for

transport infrastructure (not restricted to roads)• Appropriate hierarchical structuring of the

road network and clear lines of responsibilityfor provision and maintenance of each cat-egory

• Rigorous appraisal of investments in addi-tional road capacity to take into account (a)the effects of induced traffic on benefits, (b)the benefits and disbenefits of alternatives toNMT, and (c) environmental impacts

• Identifying the most important person orfreight movements (not necessarily vehiclemovements) both for the economic healthof the city and the welfare of the poor andconcentrating road improvements to helpthose movements.

For road maintenance• Introduction of maintenance management

systems• Funding for adequate maintenance of the

economically sustainable road capacity• Improvement of private sector contracting

capability to improve maintenance efficiency.

For traffic management• Management of traffic to maintain safe, effi-

cient, and environmentally acceptable move-ment of people (and not just of vehicles)

• Prioritization of infrastructure managementand use to protect (a) the economic effi-ciency of the city and (b) movements ofpublic and NMT, against unrestricted expan-sion of private motorized transport.

For demand management• Management of demand for road space to

balance movement with capacity throughrestraint measures

• Full-cost charging for road use (see chapter 10).• Integration of financing to recognize the

interaction between roads and other modesof urban transport (see chapter 10)

• Development of agencies capable of man-aging and maintaining the urban roadsystem.

NOTES

1. Heggie and Vickers 1998.2. Robinson, Danielsson, and Snaith 1998.3. A fuller review of available instruments of

traffic control for developing countries can befound in Cracknell 2000.

4. NEA Transport Research and Training 2000.5. We must bear in mind the need to provide

adequate grade-separated crossings to preventarterial roads from damaging the social integrityof established settlements.

6. Distributionally weighted values of time havebeen considered in some Bank projects (for exam-ple, in road projects in Vietnam), but there remainserious problems in the development of any formalprocedure for assigning distributional weights.Where the evaluation is based on traffic data mod-eled at a fairly fine zonal level, it is possible toapproximately identify the distribution of benefitsby income group (as piloted in the appraisal ofsome public transport investments in Brazil). Thatis not difficult technically and should become anormal output of economic evaluations.

As cities increase in size to the point at whichwalking can no longer satisfy the major triprequirements of the citizens, public transport—together with bicycling—becomes the majormode of transport for the poor. Buses are themain mechanized mode, carrying 6.5 trillion (6.5x 1012) passenger kilometers per year in 3 mil-lion vehicles, of which over 2 million vehiclesoperate in cities. In addition there are over 2million paratransit vehicles operating in thesecities.1

Public transport should not be viewed as onlyfor the poor, however, as the importance ofpublic transport to all income groups in manyrich European cities demonstrates. Improvingefficiency in public transport must be concernednot only with keeping costs down but also withproviding a flexible framework within which theless poor, as well as the poor, can use publictransport with confidence and comfort. If ade-quate public transport is not available, then therich will use private automobiles while the rela-tively poor will shift first to bicycles, then tomotorcycles (Vietnam and Indonesia), then totaxis (China and Indonesia), and ultimately toinexpensive cars as their incomes increase. Thefailure of conventional public transport may also

generate a burgeoning small-vehicle paratran-sit sector that can contribute to maintainingaccessibility, but that may have adverse conse-quences for congestion, air quality (pollution),and urban structure.

THE URBAN BUS SECTOR

In many developing countries of Africa, Asia, andLatin America, bus services were at one time pro-vided by regulated monopolies. In colonialregimes these monopolies were often ownedand managed by expatriates as subsidiaries ofmajor suppliers in the colonizing country. In thepostcolonial period, they were taken over intonational ownership but continued to operate asprotec ted monopol ies . In the soc ia l i s teconomies, nationally owned public sectormonopolies were also the rule. In both situa-tions—former colony and former socialist econ-omy—the traditional monopolies have nowmostly collapsed.2 In some Latin American coun-tries, they have been replaced by smaller, pri-vately owned companies operating underpermissions granted by the municipal authori-ties.3 In Africa they have largely been replaced bya fragmented small-vehicle paratransit sector,

PUBLIC ROAD PASSENGER TRANSPORT 93

PUBLIC ROAD PASSENGER TRANSPORT7Public road passenger transport is a key element of a strategy to contain

congestion and environmental air pollution, as well as being essential to the poor.

When it is appropriately regulated, competition best guarantees efficient supply,

and through franchises and concessions can mobilize low-cost operations to

provide the best quality of service and price for any budget capability. Without

adequate regulation, however, competition can have some very damaging

effects. The informal sector can also contribute effectively to satisfy demand in

competitive markets.

94 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

while in Eastern Europe and central Asia, a sim-ilar process of decline is at various stages of com-pletion.4 Only in China—where operationsremain in public ownership but are adoptingincreasingly commercial approaches to busi-ness—and in a few major cities in India andEastern Europe do traditional public operatorsstill dominate.

Although the details of history vary from countryto country, the processes of public transportdecline have much in common. In many casesgovernments have attempted to use the publictransport industry as an instrument of social policyby simultaneously constraining fare levels andstructures, and by guaranteeing favorable wagesand working conditions to employees. As deficitsmount, and in the absence of a secure fiscal basisfor subsidy, first maintenance, then service relia-bility, and finally operating capacity disappear.5 Inthe process of decline, public subsidy tends tobe progressively captured by favored, but notnecessarily very poor, groups (for example, union-i zed l abor o r m idd le -c l a s s s tudents ) . 6

Overregulation also tends to discourage marketresponsiveness.

The decline is not solely a result of cash starva-tion. Public sector operations generally lackproper incentives for, or are constrained politi-cally against, acting efficiently. This can bedemonstrated by comparisons between public

and private operators in the same country, asexemplified for Delhi, India, in table 7.1.7 Thesignificance of staffing ratios in these compar-isons is a delicate issue. Clearly, if the opportu-nity cost of labor is very low because of highunemployment or low productivity throughoutthe economy, it may be quite economic to oper-ate with high staff-to-output ratios. Indeed thatis the basis on which small vehicles that are oper-ated informally are so competitive in many low-income economies. But the institutionalizedprotection of labor in a public sector bus indus-try imposes the costs of maintaining the incomeof what is often a relatively privileged group(unionized labor) on a relatively poor group(public transport passengers), as well as entrench-ing a long-term disincentive to productivityimprovement. Maintaining an artificially highlabor complement in public sector bus opera-tions is thus likely to be a very poorly targetedform of poverty reduction strategy. It is for thisreason that competition is preferred.

COMPETITION TO MAKE PUBLIC SECTOR OPERATIONS MORE EFFICIENTCompetitive pressures can be introduced in var-ious forms, both within the traditional monopolyand between firms either “for the market” or “inthe market.” Given the inherent defects of thetraditional uncontested monopoly and thedemonstrated potential of competition to gen-

DTC

83%

216

751

9.6

Measure

Peak-period fleet utilization

Kilometers per bus per day

Passengers per bus per day

Staff per bus

Private

93%

246

1,584

4.6

Note: DTC = Delhi Transport Corporation.Source: Authors.

TABLE 7.1 EFFICIENCY OF BUS OPERATIONS IN DELHI, 1995

PUBLIC ROAD PASSENGER TRANSPORT 95

erate cost reductions and service quality improve-ments, the critical issue is how to establish thebest ways of organizing competition in order tosecure the city’s strategic objectives for its trans-port system.

Many public sector transport operators areengaged in competitive procurement of equip-ment and a range of support services (cleaning,catering, professional services, construction,maintenance, engineering, and so on) to reducecosts, improve product quality, even out internalworkloads, and eliminate the need for peakcapacity. It is good management practice to con-duct regular assessments in order to comparethe cost of undertaking functions in-house withthat of subcontracting to outsiders. The combi-nation of some freedom to subcontract with per-formance agreements between the operatingagency and its political master is one way ofattempting to improve performance. Competitionin performance can also take place between unitsperforming similar functions within an organiza-tion or by benchmarking on bus operators inother cities or countries. However, such arrange-ments tend to offer only weak incentives to man-agement, poor leverage over factor suppliers(particularly labor), and to be poorly enforced.8

COMPETITION FOR THE MARKETFirms can compete for the market in several ways.

• Gross cost service contracting involves theprocurement by a public authority from anoperator of specified services at a price deter-mined through competitive tendering.Contracts are usually for three to five years.The operator passes all on-bus revenues tothe procuring authority and does not takeany revenue risk. This system requires asecure means of ensuring that the procuringauthority actually gets any fares that are paidon the vehicle, and careful monitoring toensure that suppliers actually do provide theservice for which they have been contracted.

• Net cost service contracting is similar togross cost contracting, except that the oper-

ator keeps the revenue and hence incursboth the revenue and supply-cost risks. Thisincreases the incentive to the supplier to pro-vide the service contracted for (otherwise heloses his fare revenue)9 and obviates theneed for complex fare collection and secu-rity arrangements. However, it makes modalcoordination more difficult and often involveshigher net cost for the authorities, since thesupplier is incurring an extra revenue risk,against which he is averse, and for which hewill require remuneration.

• Management contracting involves operatorresponsibility for the management of asystem’s operation, possibly including serv-ice specification, within agreed-on parame-ters. The customer authority usually ownsoperational assets, although the operatormay be responsible for their procurementand maintenance as well as negotiating laborwages and conditions. Intermodal coordi-nation is relatively easy to achieve with thisdevice. As long as the payment arrange-ments are well structured, there is also a highincentive to provide high-quality service toattract customers. The weakness is that thecompetitive pressure may be fairly weak,trade union power relatively strong, andcosts relatively high.

• Franchising involves the grant of an exclu-sive right to provide a service that meets anumber of general quantity, quality, and pricestandards established by the authority, usu-ally because of a competition. The franchisemay be for a self-contained area, such as atown or sector of a larger city, but it is alsopossible to have route franchises—especiallywith fixed track systems. They differ fromservice contracts in allowing the contractora greater degree of freedom to develop thesystem. The franchisee may have to be paidby the authority to provide service and farecombinations that are not commerciallyviable.

• Concessions involve the granting of an exclu-sive right to provide a service without pay-ment by the authority, although the authority

96 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

may attach conditions, such as maximumfares or minimum service requirements. Inall other respects the concessionaire is actingon his own behalf and not as an agent of theauthority. Contracts are usually for ratherlonger periods, often 10 years or more, toallow the contractor to benefit from hisdevelopment of the market.

The relationship between these various forms isshown in figure 7.1. There is a rapidly develop-ing body of experience with these competitiveforms resulting largely from the wave of regula-tory reforms in Western Europe. Competitive ten-dering of service rights has also begun to extendinto the developing and transitional economies.10

COMPETITION IN THE MARKETThe most direct form of competition is that of atotally open market, in which there are no restric-tions on transport operators except thoseimposed by general law on business practices,vehicle construction and use, vehicle emissions,and highway and traffic matters. Even where thereis no quantitative limitation on competition, theopen market is usually associated with some formof quality licensing, which specifies minimumconditions for entry, including vehicle specifica-tions, environmental performance, and mainte-nance standards. In some cases the qualitativeconditions may also cover the type of service tobe operated (including stopping places), fares,and trading practices.

A more restricted form of competition in themarket may occur where, although there may beseveral operators providing services in competi-tion with each other, the total number of vehiclesallowed to operate is limited by the authority.This is a very common form of regulation for taximarkets. Particularly where fares are also con-trolled, this usually results in licenses acquiringvalue as a “business asset.”

Competition in the market gives suppliers thegreatest degree of freedom to respond to con-sumer demand and gives to the consumer the

most direct instrument—his willingness to pay—to influence what is supplied. But market com-petition is not responsive to several importanttypes of “market failure.” First, if there is insuf-ficient demand to meet the costs of supply, thenthere will be no service, irrespective of theimportance which society attaches to the pro-vision of some basic minimum service level.Second, the market is not responsive to variousexternal effects, such as congestion and envi-ronmental impact, unless they are directlycharged for. Third, because of information asym-metry and the difficulties associated with “shop-ping around,” the process of competition mayresult in a combination of price and quality ofservice supplied, which is not what the majorityof consumers would prefer. Fourth, it will not bein the interest of the individual bus operators toadapt their services and fares to promote modalintegration.

These drawbacks are not merely theoretical.Deregulation of public transport in Santiago, Chile,in 1988 resulted in massive overprovision of capac-ity, increased urban congestion, and environmentaldegradation, as old and unsuitable vehicles wereintroduced into service, and large increases in faresoccurred as operators responded to declining loadfactors. A number of aspects of anticompetitiveor antisocial on-the-road behavior have alsooccurred in deregulated or inadequately super-vised markets, including:

• “Hanging back,” to maximize patronageeither on the road or at terminals (Accra)

• “Blocking,” to obstruct rival operators’ serv-ices (Santiago during deregulation)

• “Racing,” to beat rivals’ vehicles in pickingup passengers (Nairobi and Bogotá)

• “Turning back,” when lightly loaded, to pickup passengers waiting to travel in the oppo-site direction (Pusan, Republic of Korea).11

CHOOSING AN APPROPRIATE SYSTEMFOR MANAGING COMPETITIONWhere well-managed competitive regimes havereplaced public sector monopolies in cities in

PUBLIC ROAD PASSENGER TRANSPORT 97

FIG

UR

E 7

.1

UR

BA

N P

UB

LIC

TR

AN

SP

OR

T C

OM

PE

TIT

ION

: R

EG

IME

S A

ND

TR

AN

SIT

ION

S

?

?

PU

BLI

CM

ON

OP

OLY

RA

ILD

OM

AIN

GR

OS

SC

OS

TS

ER

VIC

EC

ON

TRA

CT

ING

MA

NA

GE

MEN

TC

ON

TRA

CT

ING

NE

TC

OS

TS

ER

VIC

EC

ON

TRA

CT

ING

FR

AN

CH

ISE

S

CO

NC

ES

SIO

NS Q

UA

NT

ITY

LIC

EN

SIN

G

QU

ALI

TY

LIC

EN

SIN

G OP

EN

MA

RK

ET

LESS

RE

GU

LA

TIO

N

PA

RA

TRA

NSI

TD

OM

AIN

BU

SD

OM

AIN

—W

HO

LEF

IEL

D

L E S S P U B L I C F U N D I N G

Fina

ncia

l dif

ficu

ltie

s an

d

Op

erat

ing

effi

cien

cies

Op

erat

ing

effi

cien

cies

wit

hst

rong

reg

ulat

ion

+ p

riva

tiza

tio

n

Ret

ain

asse

ts +

pri

vate

ski

lls/e

xper

tise

pri

vati

zati

on

serv

ice

ince

ntiv

es Red

uce

auth

ori

tyin

volv

emen

t

Pri

vati

zati

on

of

asse

ts

Ban

krup

tcy/

po

or

per

form

ance

Red

uce

auth

ori

ty r

equi

rem

ent

and

fina

ncia

l sup

po

rt

Pub

lic a

cqui

siti

on

of

ow

ners

hip

to

imp

rove

ser

vice

s

Allo

w “

on

the

road

” co

mp

etit

ion

Lim

it c

om

pet

itio

n w

ith

oth

er s

ervi

ces

and

/or

red

uce

cong

esti

on/

ove

r su

pp

ly

Form

atio

n o

f o

per

ato

ras

soci

atio

ns

Con

solid

atio

nof

ope

rato

rs

Fina

ncia

l dif

ficu

ltie

s an

d

sub

stan

tial

op

erat

ors

ava

ilab

le

no s

ubst

anti

al o

per

ato

rs a

vaila

ble

Fina

ncia

l or

inst

itut

iona

l co

llap

se

Sour

ce: H

alcr

ow

Fo

x 20

00b

.

98 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

industrialized countries—such as London (UnitedKingdom), Stockholm (Sweden), and Copenhagen(Denmark)—costs per unit of output have fallenbetween 20 and 40 percent,12 and service levelshave been maintained (figure 7.2).

The policy message is clear. Well-managed com-petition can be of great benefit to the poor, butbadly regulated competition can have some verydamaging consequences. Because of this, it iscrucial to choose a competitive regime appro-priate to the objectives of the procuring author-ity, the nature of the system being managed(particularly its size and number of modes), thepotential strength of competition in the supplymarket, and the administrative capability of theprocuring authority (“getting the right frame-work”). It is also crucial to make sure that thegeneric system is well adapted to the local cir-cumstances and that it is well managed and reg-ulated (“getting the framework right”).

GETTING THE RIGHT FRAMEWORKThe objectives of the authority are the first con-cern. Both the achievement of multimodal coor-dination and the implementation of distributionallymotivated subsidy structures are easier to achievewith a small number of suppliers (concessioningor area franchising rather than route contracting),and when the supplier is not dependent on directfare revenue (gross rather than net cost contracts).On the other hand, costs are likely to be lowestwhere competitive pressures are strongest (withshorter contract-based route systems).

The larger the system and the greater the numberof modes involved, the more complex will be thecoordination problems. If the authority itself doesnot have the administrative skills to perform thisfunction, then it may best obtain that servicethrough a system concession with an experiencedspecialist company. Many French cities haveeither management contracts or system fran-chises for this reason.13

It is also clear that it is easier to operate a com-petitive system when there are already several

suppliers to the local market of appropriate sizeand competence. This situation is, however, sus-ceptible to change. If there is only one incum-bent public sector monopolist, it can be split intoseveral smaller competing units, as in London. Ifthe operators are too numerous and too frag-mented, they can be combined into a smallernumber of groups, as with the “empresas” inBogotá, Colombia, or the operators’ associationsmore recently developed in the cit ies ofUzbekistan.

GETTING THE FRAMEWORK RIGHT Whichever system is chosen, effective competi-tion between private sector suppliers can onlybe achieved if the public sector itself is appro-priately structured and capable. This imposes anumber of critical institutional requirements:

• Political supervision of public transport thatis separated from professional management

• Service planning that is separated from serv-ice provision, and adequately staffed andskilled

• Acquisition of new procurement skills, in thecase of franchising or contracting

• Operations privatized, or at the very leastcommercialized

• Public company operation units restructuredin a form conducive to competition, or sub-ject to strong external competition.

This program of reforms may take time andrequire progressive refinement, especially whereit involves concepts and procedures that are novelto the country (box 7.1).

Contracts must be of clearly defined duration.For route service contracts where the procuringauthority is defining fare and service levels, thecontracts can be of relatively short duration (threeto five years). Particularly where there is a regularstream of contracts coming up for bid, it is notnecessary for the contract length to reflect buslife, since vehicles can be switched between con-tracts either through secondhand markets orthrough leasing arrangements. Extension of con-

PUBLIC ROAD PASSENGER TRANSPORT 99

tracts saves tendering costs but can blunt com-petition and, where it is allowed to become thenorm, can be the basis on which an ostensiblycompetitive system becomes captured by a cartelof existing operators.14

Contracts must also define the rights and dutiesof the parties in as complete and consistent away as possible. If fares are controlled, contractsshould define the process for their adjustmentto account for general cost inflation, as well as

FIGURE 7.2 EFFECTS OF COMPETITION ON BUS TRANSPORT IN LONDON

0

100

200

300

400

500

600

444

270

15

0

5

10

20

25

30

35

0

5

10

20

25

3029

22

0

50

100

150

200

250

300 263 241

0

5

10

15

20

25

17

20

8

4

12

0

2

6

10

14

4

12

0

2

6

10

14

0

2

6

10

14 12

2

0

50

100

150

200

250

182

29

PENCE PENCE

PENCE PENCE

PENCE PENCE

COST PER BUS MILE (–40%) COST PER PASSENGER MILE (–22%)

REVENUE PER PASSENGER MILE (+19%)REVENUE PER BUS MILE (–8%)

GRANT PER BUS MILE (–84%) GRANT PER PASSENGER MILE (–80%)

1994–95 (estimated outturn)1984–85

Source: Transport for London (TFL) figures at 1994 prices.

100 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

define the compensation for any discretionaryfare adjustments introduced by the procuringauthority. If this is not properly provided for, fran-chising systems are doomed to failure, asoccurred in Jamaica.15

It is possible, in principle, for competition to oper-ate between privately and publicly owned oper-ators, but that type of operation can only workeffectively if the public sector operators are strictlycommercialized, subject to a bankruptcy con-straint on their commercial behavior, free fromspecific public service requirements not imposedon private competitors, and not eligible to be“bailed out” directly or indirectly by central orlocal government. Such operation will probablyrequire some legal change in the operators’status, and will probably only be secure if there isalso an independent auditing arrangement toensure that the operators do not bid below costs

to obtain business. These were the conditionsunder which public and private sectors competedin London for the interim period before the publicsector operations were privatized.

A phased reform is possible as long as there is asufficiently clear program and timetable to giveprivate competitors the confidence that the reformwill be consummated, and to give the publicsector operators an incentive to adjust in prepa-ration for competition and privatization, ratherthan to dig in politically to prevent it. In the caseof the reform in London just mentioned, therewas a timetable both for the extension of com-petitive tendering through the whole network andfor the reorganization and privatization of thepublic sector operators. There will inevitably bea tendency to argue that subsidies should bephased out slowly in order to avoid any adverseimpact on fares. In practice, most of the benefits

BOX 7.1 INTRODUCING COMPETITIVELY TENDERED FRANCHISES IN UZBEKISTAN

Urban public transport services were traditionally supplied in Uzbekistan by state-owned enter-prises that enjoyed areawide, sometimes citywide, monopolies. Beginning in late 1997, however,as part of the transformation of this former socialist economy into one that functions on marketprinciples, the Uzbek government implemented radical changes in the organization and regula-tion of urban public bus transport services. Through a gradual and carefully planned process,which included experiments in a few cities, a study tour to London, and progressive scaling upto all secondary cities, responsibility has been given to the city administrations to organize all busservices on the basis of exclusive route franchises. These franchises are allocated through a com-petitive tendering process open freely to private companies and associations of small owner-operators as well as the state-owned enterprises. Tendering is under the responsibility of a specialcommission in each city, chaired by a deputy mayor, and operating under precise rules set by atransport regulatory agency in the central government. Bidders’ discounts—if any—from the pas-senger fare ceiling, proposed service frequency, and bus fleet characteristics are the main selec-tion criteria. Franchise duration, initially set at six months, renewable once for another six months,is progressively being extended (and is now at one year).

These reforms, completed in two years, have resulted in impressive changes. Numerous privateoperators have entered the public transport market, many new jobs have been created in theemerging bus service sector, and a healthy competition has developed (particularly for the rap-idly growing minibus services). Private operators now supply more than 50 percent of all urbantransport services. A bus route franchising system is also now being implemented in Tashkent.

Source: J-C. Crochet, from World Bank project files.

PUBLIC ROAD PASSENGER TRANSPORT 101

are likely to arise in the initial round of tenders,so it is advisable that contracts be let from theoutset on conditions that are likely to be finan-cially sustainable in the long run.

PARATRANSITOne of the most notable features of the publictransport sector in the developing and transi-tional economies in recent years has been theexplosive growth of publicly available passengertransport services outside the traditional publictransport regulatory system, often referred to asparatransit.16

A number of characteristics are typical of para-transit services, although not necessarily appli-cable in all cases. These include:

• Services are usually unscheduled and often,though not always, on demand-responsiveroutes, filling gaps in formal transit provision.17

• The vehicles operated are typically small,including motorcycles,18 partly because ofthe greater ease of financing and flexibilityof operation of the small vehicles, and partlybecause controls over small vehicles are laxeven in situations where entry to the large-vehicle market is strictly controlled. In somecities, such as Damascus, Syrian ArabRepublic, small vehicles dominate the market.

• The vehicles used are often old, having beenretired from other countries or other usesdomestically, so that the capital investmentnecessary to enter the business may be small.

• The vehicles used are also often very simple,including, in many countries, NMT vehicles.Some of these vehicles, such as the motor-ized rickshaws of East and South Asia andthe jeepneys of Manila, are very specialized,but in many cases they are simply adapta-tions for passenger carriage of whatevervehicle is inexpensively available (includingconverted trucks in Africa and motorcycletaxis in Bangkok).

Paratransit services are usually provided by infor-mal operators with the following characteristics:

• They are “noncorporate,” usually operatingas single-person enterprises, although fre-quently with a vehicle owner who is not theoperator. Often the driver pays a daily feeto the owner, incurs operating and mainte-nance costs, and keeps all revenue in excessof the fee. This gives a high incentive to worklong hours and to obtain paying passengersby all available means, including touting,poaching, racing, and so on.

• They are often outside the tax system or ben-efit from favorable treatment of the non-corporate sector.19 They may also have anadvantage in competition with public sectoroperators, with costs inflated by minimumwage regulations, strict working hourrequirements, neglect, and corruption.

Paratransit performs many roles. In Africa it is thedominant mode of public transport of the poor.In the former Soviet Union, it supplements adeclining formal sector. In East and South Asia,and to some extent in Latin America, it comple-ments the formal sector, providing differentiatedservices in identified market niches. In other partsof Latin America, it increasingly competes head-on with the traditional suppliers. Paratransit pro-vides a range of services including:

• Feeder services linking inaccessible hous-ing areas to the main transport routes (thefour-wheel drives in the barrios of Caracasor the cycle rickshaws of Dhaka)

• Local distribution in inaccessible areas thatare not served, or are underserved, by con-ventional public transport (Lima)

• Trunk services complementing, or compet-ing by quality differentiation with, the formalsector on major routes (the minibuses ofmany central Asian countries or the “peruas”[passenger vans] of São Paulo and manyother Brazilian cities)

• Direct longer-distance services on routeswhere the formal sector supply is slower orinfrequent (the “truchos” [trunk routes] ofBuenos Aires)

• Duplication of franchised services.

102 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

There are many different combinations of infor-mal transport structure, organization, service, andvehicle type. Table 7.2 classifies the examplesmentioned above.20

The problems of informalityThe informal sector is often viewed as a nuisanceby national and municipal transport authorities,particularly when those authorities are responsi-ble for the provision of conventional bus serv-ices. Despite this view, the sector has someimportant merits that have led the internationalinstitutions, including the World Bank, to view itmore favorably, not least as a source of employ-ment for the poor. In many countries it representsa very significant entry point to urban employ-ment. For example, pedicab drivers in many Asiancountries have some of the longest hours of work(typically 70 per week), lowest levels of educa-tion, and lowest incomes of all categories of work-ers, and include a disproportionately large shareof recent rural-to-urban migrants. In many Asian

cities, it is estimated that over 15 percent of thepopulation is dependent directly or indirectly oninformal sector transport for their livelihood. InDhaka, the proportion has been estimated atover 25 percent.21

The services that the informal transport sectorprovides are also valuable. Particularly in SouthAsia, informal transport performs a feeder func-tion for relatively well-off people. In other cases,particularly in lower-income countries in Africa,it is often the mode of transport “of the poor”as well as “by the poor.” In some cites it may doboth—for example, in Manila 45 percent of tripsare carried by the “down-market” jeepneys butanother 12 percent by the “up-market” FX mini-vans. It is usually very market responsive, pro-viding access to poor areas, direct routing, speed,and flexibility of service. If there is a demand forthese characteristics that is not being met by theformal sector, the informal sector will invariablymeet it if permitted (and often even if not per-

TABLE 7.2 A CLASSIFICATION OF INFORMAL (NONCORPORATE) URBAN TRANSPORT OPERATIONS

Vehicle Service features Passenger Service Market type Routes Schedules capacity niche regime Examples

Large bus Fixed Fixed 25–60 Line-haul Franchised Buenos Aires; Rostov, Russian Federation

Minibus Fixed Fixed/ 12–24 Line-haul Franchised São Paulo; Bangkok; Harare, Zim-semifixed babwe; Johannesburg, South Africa

Jeepney Fixed Semifixed 12–24 Line-haul Franchised Manila

Microbus and Fixed Semifixed 4–11 Feeder Licensed Caracaspick-up truck

Shared taxi Variable Variable 3–6 Short trips Licensed Casablanca, Morocco; Lima; Mara-caibo, Rep. Bol. de Venezuela

Three-wheeler Variable Variable 2–4 Short trips, Unregulated Phnom Penh, Cambodia; Delhi; feeder Bangkok; Jakarta

Motorcycle Variable Variable 1–4 Feeder, Unregulated Bangkok; Cotonou, Benin; some Lomé, Togo; Douala,longer Cameroon

distances

Pedicab and Variable Variable 1–6 Short trips, Unregulated Dhaka; Vientiane, Lao People’shorse-drawn feeder Dem. Rep.; Mumbaicart

Notes: “Franchised” means holding official permission specifying task, area of operations, and so on. “Licensed” meansholding unspecified permission to operate the vehicle.Source: Based on Cervero 2001.

PUBLIC ROAD PASSENGER TRANSPORT 103

mitted!). This high degree of market respon-siveness means that there may be little need forgovernment support or economic regulation. It isinherently fragmented and hence highly com-petitive, although that has disadvantages as wellas advantages, and typically results in the emer-gence of either formal regulation or informal self-regulation through operators’ associations.

Despite these advantages, informal transport hasa very poor image and reputation. It is often avery low earnings sector, with crews exploited byvehicle owners. It has an association with poverty,viewed as symbolically inappropriate by gov-ernment, which tries to reduce its role as muchas possible. There appear to be three mainaspects of informal transport systems that con-tribute to this negative image:

• Dangerous on-the-road behavior and asso-ciation with crime and violence

• Urban congestion and adverse environ-mental impacts resulting from use of small,old, and ill-adapted vehicles

• Undermining of basic network of existingservices.

These defects are frequently exploited by vestedinterests. Police and other public officials maytake advantage of the quasi-legal nature of thesector to supplement their incomes.22 Traditionaloperators also exploit the limitations of the infor-mal sector as reason for protecting the formalsector. The policy quandary is how to distinguishbetween real problems and the special pleadingof vested interests.

Controlling operating practicesParatransit is often criticized because the opera-tor will only provide service where he considersit worthwhile to do so. But in many circumstancesthat will not matter. In particular, where the driveris leasing the vehicle on a daily basis, he may onlybe able to make a surplus over the rent by start-ing early or finishing late, or both, and plying histrade wherever there are passengers to be found.His standard of “remunerativeness,” or prof-

itability, is thus very low, and his service cover-age prolific. As experience has proved over manyyears in Buenos Aires, and more recently in thesecondary cities of Uzbekistan, the organizationof informal operators into route associations canensure disciplined service.

Where there is a desire to provide even moreservices than the informal sector can providecommercially, it is argued that a monopolist publicsupplier is required. That is, of course, a falla-cious argument. Directly subsidized services canbe efficiently obtained through competitive ten-dering of franchises. Even cross-subsidy can beorganized within a competitively tendered fran-chising system—as in London and several otherlarge cities in Europe—either by packaging prof-itable and nonprofitable services together in ten-dered lots or by using fees from “positive”concessions to finance “negative” ones.

The most common concern about a fragmentedinformal sector is that the competitive pressure toearn a living will result in excess capacity, low loadfactors, and antisocial and often dangerous oper-ating practices, such as lack of attention to pas-senger safety, racing, turning short, blockingintersections while touting for traffic, and so on.Certainly there is evidence of such behavior in anumber of cities, including Kingston, Jamaica,and Harare.

Exponents of free markets have frequently arguedthat, in the long term, operators will see that it isnot in their own interest to continue such unde-sirable practices. Typically this results in the for-mation of associations that limit entry andorganize more disciplined service. Such associ-ations are the norm in most of Sub-Saharan Africaand are common wherever the informal sector isunregulated.

There are several problems with such self-regula-tion. First, because it is outside public control, theassociation acts in the interests of its membersand suppliers, and not in the interests of its cus-tomers. During the initial period of complete

104 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

deregulation in Santiago, Chile, the action of theoperators’ cartels led to a rapid increase in fares.23

Second, because self-regulation is not based onany legal rights of exclusion, it is often enforcedby violent means, as occurred in the taxi (minibus)sector in South Africa. Third, the need to ensurefair allocation of revenues between members oftenresults in suboptimal operating practices. In par-ticular, ensuring that all vehicles are dispatchedfrom the terminal with full loads equalizes incomesat the expense of passengers (forcing them to walklong distances to terminals to access the service)and utilization of vehicles (forcing delays as theyqueue for their turn to depart). Only the moresecure and long-standing associations are able toadopt more efficient practices—and that extrasecurity may involve more monopoly power.

Congestion and environmental impacts The pressure of competition may also lead to anexcess supply of vehicles (in the sense that morevehicles are in service than are necessary to provideuncrowded service at high frequency) and the useof small and often inexpensive older vehicles. Smallvehicles are usually much simpler and lighter inconstruction than are conventional buses. Becauseof this, both capital and operating costs per vehi-cle seat vary relatively little with respect to vehiclesize. If labor costs are also low, there is no incen-tive to use large vehicles, which the informal sectorwould find difficult to finance in any case. Becausethe effects of congestion and environmentalimpact are external to the individual operator, themain incentive is to operate inexpensive, andhence often older, vehicles. The result is that totallyunregulated entry in low-income countries is likelyto result in a higher level of congestion and envi-ronmental impact than is socially desirable.

Two economic distortions have contributed tothe explosion of informal services in small vehi-cles. First, there is often an excess supply of laborin urban areas that coexists with minimum publicsector wage rates and inefficient operation forthe formal operators. Second, in the absence ofany pricing system for the use of scarce roadspace or adequate proxy priority given to large

vehicles, the informal sector small vehicle is ableto provide a faster, and sometimes less expen-sive, service than is the formal operator.

Undermining of basic network of servicesA subtler problem has been emerging in recentyears. In many cities in Latin America (such asBuenos Aires, São Paulo, and Fortaleza) and insome cities in East Asia (such as Bangkok), infor-mal operators are beginning to operate servicesin direct competition with traditional large-vehi-cle services, whether operated by public or pri-vate enterprises. The basis for informal operators’ability to compete in this way has often been that,by operating smaller vehicles and a denser net-work of services, they are able to offer a quicker,more convenient door-to-door service than canthe traditional operator. In some cases (as typi-cal in Brazil), this service is provided at fares (andsometimes on routes) identical with those of thetraditional operator.24 In other cases the servicemay be operated at a premium fare. In eithercase, the effect may be to reduce the demandfor the services of the traditional operator andthus either increase the breakeven fare or reducethe breakeven frequency. Both responses wouldbe to the disadvantage of those passengers cap-tive to the traditional services.

Probably the most serious impact is that on thedevelopment of integrated multimodal serviceand fare structures. In a number of Brazilian cities,where metros or suburban railways have beenrehabilitated and bus networks restructured inthe context of an integrated fare system, pas-sengers are being lost to informal operators whoprovide direct service at competitive fares. Thecritical question is what, if anything, public author-ities should do to respond to or control thismarket-oriented response.

THE FUTURE OF PARATRANSIT ANDTHE INFORMAL SECTOR Given the importance of paratransit both as anincome generator and, often, as a service providerto the poor, attempting to eliminate it by admin-

PUBLIC ROAD PASSENGER TRANSPORT 105

istrative action could generate significant unrest.Following action to control the sector in São Pauloin 1999, roads were blocked and 24 formal sectorlarge vehicles were destroyed in a period of threemonths. Repression is thus not a likely solution tothe perceived problems. Rather, governmentsshould examine why the informal sector exists,and then try to identify a regulatory and admin-istrative framework within which the potential ofthe sector can be mobilized and developed.

Many of the defects attributed to the sector canbe attributed to its insecurity. Predatory behavioron the road is necessary to make a living in a con-text of very low opportunity costs and, hence, inthe context of prol i ferat ion of capacity.Inadequate capitalization, and the consequentsmall size and poor quality of the vehicles used inmany cases, may in turn be attributed to theabsence of a sufficiently secure expectation ofthe future revenue to justify commitment of cap-ital to large assets that lack versatility.

A number of different approaches have beenadopted to overcome this lack of a secure field ofoperation. Several countries allow free access incertain specialized markets (local feeder busesin Seoul, air-conditioned services in Dhaka, com-muter charter bus services in Delhi), but thesetend to be limited niche markets, and oftenrequire a higher class of vehicle to attract patron-age. In São Paulo, a provision for the formal sectoroperators to accommodate 15,000 “peruas” (pas-senger vans) to supplement their own serviceshas failed to defuse an explosive situation thathas involved the suppression of three times asmany existing (albeit illegal) operations.

A rather different approach to the problem of thecore supply of informal transport services is thecreation of “curb rights,” permitting registeredinformal sector operators to pick up and set downpassengers in specific areas, but not otherwise con-straining their activities. The aim is to give a super-vising authority some leverage (withdrawal ofregistration) to discourage antisocial on-roadbehavior, while leaving freedom to the operators to

respond flexibly to demand. It is very similar to thelicensing arrangement found in rank-based taximarkets in many industrialized countries, but it isusually supplemented with some control on faresor capacity and also often overlaid with regulationor self-regulation to determine access priority. Ithas not been applied on any substantial scale tobuses or minibuses, and would probably be verydifficult to enforce in developing countries.

The more common solution in the bus sector hasbeen found in the form of medium-term routefranchise contracts. The immediate impedimentto the inclusion of the informal sector in such asystem is often the desire of the municipal author-ities to guarantee regular, scheduled service onroutes requiring a large number of vehicles. Thisimpediment can be overcome by combining fran-chising (preferably competitively tendered) withfreedom of establishment for (and indeed someencouragement to) operators’ associations. Thatsolution was the basis on which the urban bussector in Buenos Aires operated very effectivelyuntil it began to be undermined by a new influxof illegal shared-taxi operations. It is also the basison which competitively tendered franchising isbeing introduced in countries of the former SovietUnion, such as Uzbekistan and the Kyrgyz Republic.Some 2,700 informal sector vans have recentlybeen legalized and regulated to give alternativeservices as cooperatives in Rio de Janeiro.

The main problem in pursuing that regulatorypath is to determine how best to prevent collu-sion and the emergence of a grand cartel able toexploit monopoly power. In Argentina this wasachieved, despite the absence of competitive ten-dering, by ensuring that route franchises grantedto specific associations overlapped, so that therewas a degree of competition on the road. InUzbekistan it has been done by official encour-agement for the creation of multiple associations.

Competitively tendered franchising arrangementsalso make it possible to address the issues of con-gestion (which can be addressed by limiting theamount of capacity franchised to operate in par-

106 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

ticularly congested streets) and environment(which can be addressed by putting qualitativestandards or criteria in the selection process). Bothproblems have been very satisfactorily addressedin Santiago Chile, although road and rail passen-ger transport remain as competing alternatives,often serving different income groups, rather thanas part of an integrated network in the absence ofany institutionally systematic provision for improv-ing modal coordination.

Some additional encouragement may also beneeded. Restricted access to credit limits the abil-ity of many operators to buy their own vehicles,forcing them into a dependence on an absenteeowner. Even with the introduction of franchising,there will be a period, until the system is well estab-lished, in which it will be difficult for operators tosecure funding for vehicles against a franchise con-tract. Assistance with vehicle finance may be a nec-essary component of reform. The ultimateobjective should not be to maintain a highly frag-mented bus industry for its own sake, but ratherto encourage more informed and disciplinedentrepreneurial structures on which competitioncan be based. In fact, most of the operators’ asso-ciations in Buenos Aires developed naturally frompools of privately owned vehicles to shareholdingcompanies. Some of the recently formed opera-tors’ associations in Uzbekistan are already begin-ning to undertake functions, such as jointpurchasing, which are steps to corporate form.

In summary, there are two conflicting considera-tions concerning the role of the informal sector.On the one hand, the services provided by theinformal sector may better respond to consumerdemand than those of the formal sector, andemployment in informal transport may be one ofthe few areas of gainful economic activity opento new rural-to-urban migrants. On the otherhand, informal transport brings with it adverseeffects on congestion, environment, and basicpublic transport network viability.

The balance of these considerations may be in favorof paratransit in smaller cities, where excessive

supply is not a problem, but against it in larger, con-gested, and polluted cities. The critical questionthen becomes at what point does the growth ofparatransit need to be controlled and redirectedeither to niche markets or more formal arrange-ments. In that progression, some policy measuresshould be specifically directed to attenuating con-gestion or environmental impacts directly wherethey occur. Limiting access to particular locationsand enforcing environmental and safety regulationsare the appropriate policy instruments—universalprohibitions of the informal sector are not. Even inlarger cities, the appropriate policy response wouldseem to be one that permits informal operators tocompete for franchises in sectors of the marketwhere their flexibility is particularly advantageousbut where their small vehicle size is acceptable.Authorities thus need to plan for the use, devel-opment, and migration of paratransit in a more pos-itive way than has hitherto been common.

CONCLUSIONS: TOWARD ASTRATEGY FOR PUBLICTRANSPORT

Public transport is critical to the welfare of theurban poor and a crucial element in any poverty-oriented city development strategy. Yet it is fail-ing to provide the necessary service and is actuallyin decline in many developing countries just atthe time when many much richer, industrializedcountries have begun to recognize its impor-tance. That decline has some technical roots butis mostly a consequence of the inappropriate-ness, for the tasks expected of it, of the institu-tional and financial arrangements under which ittypically operates.

The main elements of a strategy for urbanpublic road passenger transport are suggestedas follows:

On planning and integration• Public transport provision should be treated

as a key component of a city developmentstrategy or structure plan.

PUBLIC ROAD PASSENGER TRANSPORT 107

• Public transport must be given a high prior-ity in the design and use of scarce road space.

On competition• Planning of public transport service should

be separated from provision of public trans-port service.

• Competition should be recognized as thebest way to secure good value for money inpublic transport.

• In complex cities the best form of competi-tion for the market may be through tenderedfranchises or concessions.

• The competitive regime should be designedand regulated to maintain healthy compe-tition and avoid excessive oligopolistic ten-dencies.

• City administrations should be restructuredto facilitate competitive procurement of serv-ices (see chapter 11).

On paratransit and the informal sector• The role of paratransit in satisfying dispersed

trip patterns and in flexibly addressing thedemands of the poor should be recognized.

• Anticompetitive or antisocial behavior withinthe sector should be controlled through theestablishment and enforcement of qualitystandards.

• Cities should strive to find ways to mobilizethe initiative potential of the informal sectorthrough legalizing associations and throughstructuring franchising arrangements in orderto give the small private sector the opportu-nity to participate in competitive processes.

• Cities must ensure that informal operators meetthe same environmental, safety, and insurancerequirements as formal operators, and that they meet their proper tax obligations.

• Cities should plan for a dynamic regime thatwill allow for a transition to a more formal rolefor the informal sector when appropriate.

On pricing • General fare controls should be determined

as part of a comprehensive city transportfinancing plan, and their effect on the

expected quality and quantity of servicecarefully considered.

• Fare reductions or exemptions should befinanced on the budget of the relevant lineagency responsible for the categories ofperson affected (health, social sector, edu-cation, interior, and so on).

NOTES

1. Halcrow Fox 2000b.2. Gwilliam 2001.3. Aragão, Brasileiro, and Marar 1998.4. For further detail, see Gwilliam 2000b.5. Effective public sector operations still continue

in some developed cities, such as Vienna (Austria),Stuttgart (Germany), and Zurich (Switzerland), evenif relatively costly. The relevant ingredients for thisoutcome are efficient economies, high priority forpublic transport, and wealthy communities payingmore attention to quality than to cost. Such condi-tions are rarely found in developing countries.

6. For an example of this, see Teurnier andMandon-Adolehoume 1994.

7. Similarly, in São Paulo the now-defunct stateoperator had an average of 8.6 employees perbus compared to only 5.5 for private sector oper-ators. Even after allowing for public sectoremployment associated with activities such asplanning and coordination of the public trans-port system, labor productivity was still 28 per-cent below that of private sector operators.

8. World Bank 1995.9. There may still be some service elements

(the earliest bus, for example) that cost more torun than the revenue they earn, and for whichthere is an incentive not to supply, unless thesupply is regulated.10. See, for example, Gwilliam, Kumar, andMeakin 2000.11. This is not in itself a bad practice as long aspassengers are not seriously delayed or forcedto pay twice. In the case of fragmented compe-tition, these protections are rarely observed.12. For a more detailed discussion of alterna-tive forms of competition and some advice ontheir design, see Halcrow Fox 2000b.

108 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

13. Management contracting and system fran-chising are most common in France (Systra 2000).14. See Aragão, Brasileiro, and Marar 1998 fora discussion of how this has happened in manyBrazilian cities.15. Gwilliam 1996.16. This does not necessarily mean that they areoperating illegally, as in many countries entry tothe sector is effectively free, with operators sub-ject only to the general rules of the road and lawof the land. Nor does it necessarily mean thatthey are operating completely independently,since many informal sector operators are mem-bers of associations of operators.17. Cervero 1998.18. The motorcycle taxi is the most rapidlyexpanding segment of the market. It is esti-mated that there are 125,000 “moto-dubs” in

Phnom Penh and 100,000 “rub-jangs” inBangkok.19. For example, in some of the Central Asianrepublics of the former Soviet Union, taxation onthe informal sector amounts to less than 3 per-cent of revenue, compared with nearly 25 per-cent for the corporate sector.20. For a fuller description and classification oftypes of informal transport, see Cervero 2001.21. Gallagher 1992.22. For example, the motorcycle taxi businessin Bangkok involves buying off officialdom at sev-eral levels of the hierarchy (see Cervero 2001).Sometimes, as in Cairo, police become involvedin ownership of the vehicles.23. Dourthe and others 1998.24. Associacão Nacional de Transportes Públicos1999.

THE SCALE OF THE ISSUE

Mass rapid transit (MRT) comprises a spectrumof modes of urban public transport that use spe-cific fixed-track or exclusive and separated useof a potentially common-user road track (such asmetros, suburban railways, light rail transit, andbusways). MRT usually has superior operatingcapacity and performance compared with unseg-regated road-based public transport (such asbuses, taxis, and paratransit). Rail-based metrosystems in developing countries carry about 11billion journeys a year, surface rail about 5 billion,and light rail about 2.5 billion. While the pro-portion of public transport trips by rail is morethan 50 percent in Seoul and Moscow (and theproportion of passenger kilometers is evenhigher), rail systems dominate only in a very fewcities.

MRT can, in principle, contribute to the achieve-ment of all the main objectives of urban devel-opment policy. It can improve efficiency of thecity economy by reducing travel costs and bymaintaining a higher level of city-center activityand the associated economies of agglomerationthan would otherwise be the case. The impact ofpoverty can be directly reduced where MRT isthe major carrier of the poor, and indirectlyreduced through the benefits the poor receivefrom economic prosperity. It can also improvethe quality of life—immediately, through shifting

movements to more environmentally benignmodes of transport, and in the longer run, by sup-porting a more environmentally favorable land-use structure.

In practice, these benefits do not always accrue.Costs of rail investments are often underesti-mated and passenger flows overestimated.1

Excessive indebtedness or unrealistic calls onmunicipal resources to finance expensive MRTmodes can damage the local economy and pre-clude other socially desirable investments.Attempts to avoid that indebtedness through pri-vate financing or increased fares may disadvan-tage the poor by excluding them from use of themore expensive systems. The prohibition of par-allel bus or minibus services may sometimesincrease MRT financial viability by eliminatingless-expensive, lower-quality modes on whichthe poor were most dependent. Potential con-gestion reduction and environmental benefit maybe lost if the road space freed by a shift of pas-sengers from conventional bus to MRT is allowedto be filled up by additional automobiles.

The central challenge with respect to urban MRTis to identify the strategic objectives being soughtby the city, and then to identify means of imple-mentation which best secure the benefits andevade the disbenefits listed above. Because thereare a wide variety of city types, city objectives, MRTtechnologies, and pricing and financing mecha-

MASS RAPID TRANSIT 109

MASS RAPID TRANSIT8Mass rapid transit can contribute both to city efficiency and to the needs of the

poor in the largest cities, but it can also impose a heavy fiscal burden. Alternative

technologies should be evaluated both in operational and fiscal terms. More

expensive rail-based systems should only be adopted within an integrated

planning and financing structure ensuring system sustainability, effective

coordination of modes, and affordable provision for the poor.

110 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

nisms, there is a rich menu of strategic alternativesfrom which to choose. In many cases the problemis not simply that of exclusive choice betweentechnologies, but more that of selecting the opti-mum mix of technologies and the optimum phas-ing of MRT capacity expansion. An appropriatestrategic stance is thus not to be “for” or “against”MRT, or any particular variant of it, but to properlyappreciate the critical factors affecting choice oftechnologies, operating, financing, and owner-ship arrangements, and to ensure that the choicesmade are consistent with city characteristics, objec-tives, and economic capability.2

Rather than advocating a “one-size-fits-all” MRTstrategy, emphasis thus falls on the formulation ofcritical questions that must be addressed indesigning an MRT strategy, and the presentationof evidence and experience relevant for address-ing those questions. These critical questionsinclude:

• What are the objectives and role of MRTwithin the city development strategy?

• How should MRT relate to urban structureand land-use policy?

• How does MRT impact on the urban envi-ronment?

• What factors should be considered in thechoice of MRT technology?

• How should MRT be integrated within abroader transport sector policy?

• How should MRT be priced to gain maxi-mum benefit?

• Who should own and finance MRT?• What is the impact of MRT on the finances of

national and local governments?

OBJECTIVES AND ROLE OF MASSRAPID TRANSIT WITHIN THE CITYDEVELOPMENT STRATEGY

The reduction of road congestion, with its con-sequent economic and environmental benefits,is usually the motivating factor for investments inMRT. Both because of its capability for carryinglarge volumes and for its superiority over buses

in attracting traffic from, or limiting the trend ofdiversion to, the car, rail-based systems are usu-ally preferred for the purpose of reducing roadcongestion. In fact, however, the studies of MRToutcomes in 1990,3 and the recent update(Halcrow Fox 2000a), argue that congestion israrely reduced. Rather, the effect is to permit thecontinued development of city-center activitywhile total movement volumes on the main radiallinks, along which metros are typically aligned,increase to levels that would have produced intol-erable congestion in the absence of MRT.4 It ishence the structuring effect to avoid sprawl ofbusiness activity and the consequential economiesof agglomeration in production rather than reduc-tion of transport costs over time that is the mainsource of economic benefit.

There are several corollaries of this argument.First, if the benefits of MRT are ultimately struc-tural, then they should be planned and designedin the light of high-level objectives of a strategicstructure plan. That is rarely the case, although itis notable that some of the more successful metrodevelopments, such as that in Singapore, havebeen developed this way. Second, if MRT isdesigned to maintain the quality of access to thecenter, it needs to be supported by appropriateactions to feed the trunk links and hence needsto be developed in the context of a comprehen-sive transport plan. City-center development poli-cies should be complementary to it. Third, ifstructural impacts are the objective, that should bereflected in the way MRT investments areappraised. Finally, for service to be maintained,the development needs to be properly financed.Cutting corners to reduce costs (such as the omis-sion of elevators to high-level platforms on thenew Bangkok Transit System [BTS] and ManilaLRT3 systems) or charging high fares to maximizerevenue for private concessions may actuallyreduce the economic benefit of the investment. Afirm financial planning context is necessary to getthe best out of efficiency-oriented MRT systems.

In cases where the main function of MRT is to pro-vide basic accessibility in poor cities without alter-

MASS RAPID TRANSIT 111

native means of transport, the most importantconsideration might be the selection of a systemthat is affordable to users, or to the public budget,or to both. Where MRT is aimed primarily at reduc-ing congestion and maintaining the central city,and predominantly carries middle-class passen-gers, it should be assessed in terms of its contri-bution to the economic viability of the city (fromwhich the poor also benefit). The importantpoverty consideration in that context would beto ensure that the viability of the MRT system doesnot involve supporting policies that are damagingto the poor, and that its finance is supportablewithout damage to the maintenance of thedesired level of other basic services to the poor.

RELATIONSHIP TO URBANSTRUCTURE AND LAND USE

The integration of MRT within the urban fabricmakes some important demands on the plan-ning system. Rights-of-way must be establishedand protected. Space must be released fordepots and terminals. In addition, where high-density ancillary developments are intended, theland must be assembled into lots suitable fordevelopment and the appropriate densities ofdevelopment sanctioned.

The most indisputable structuring effect of metrosis that they allow central business districts in large,dynamic cities to continue growing, where serv-ice by road, either by car or bus, would be increas-ingly frustrated by congestion. Without thehigh-capacity links, activities would begin to bedecentralized. This has implications both for cityplanning and for project evaluation. A consciousattempt to maintain the growth of the center willsave on public infrastructure costs in other areas;avoiding these extra costs is an important partof the long-term benefit of MRT investments.Unfortunately the magnitude of those savings islittle researched, particularly in developing coun-tries, and the economic evaluation of MRT invest-ments is usually based on the more conventionaluser cost-benefit appraisal. While that may still

be justifiable in the interest of avoiding the worstkind of “white elephants,” a more wide-rangingmulticriteria analysis may be the most suitableway of ensuring that those unmeasured effectsare taken into consideration. An integrated land-use, urban transport, and air quality strategy, suchas the PITU in São Paulo, is needed to ensurethat the MRT system is adequately inserted inthe urban structure.5

Obtaining desirable structuring effects outsidethe center is more difficult. Clustered multinu-clear development associated with station loca-tions sometimes occurs spontaneously, butnormally requires either some preplanning bygovernment (as in the cases of Singapore andHong Kong, China) or close links between pri-vate ownership of the MRT system and contigu-ous developments (as is common in Japan). Inboth cases this requires land to be assembledfor development in relatively large lots. This hasbeen achieved by comprehensive public owner-ship of land in Hong Kong, China, by compulsorypublic purchase in Singapore, and throughmarket mechanisms in some Japanese privaterailway developments.6

MASS RAPID TRANSIT AND THEURBAN ENVIRONMENT

One of the most common arguments for MRTsystems is that by reducing more polluting pri-vate road traffic, they benefit the environment.Within the MRT family, rail systems are preferredto bus systems7 because their use of electric trac-tion is believed to mean that they are cleaner.8

Whether this is true, of course, depends both onthe source of electricity for the MRT and on thekind of road-based system with which the MRTis compared. Underground systems are often pre-ferred to at-grade or elevated systems, becausethey are considered less intrusive in the urbanfabric.9 Underground metros are thus regarded asthe most environmentally beneficial, even wherethe main diversion is from another public trans-port mode—the bus—rather than from the pri-

112 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

vate car. Environmental benefits were accord-ingly among those taken into account by theWorld Bank in its initial appraisal of the Line 4metro investment in São Paulo.

The main caveat to be entered against the envi-ronmental benefits of MRT concern longer-term,less-direct effects. Insofar as metro investmentsallow a higher level of activity to be sustained inthe historic city business centers than would oth-erwise be the case, the absolute level of inner-city road traff ic may be l i t t le affected incomparison with the “without metro” case.10 Thecritical question is then to what extent does itobviate growth of environmentally damaging traf-fic. Most assessments fall short of a thoroughanalysis of this counterfactual, though implyingthe net effect is beneficial.

CHOICE OF MASS RAPID TRANSITTECHNOLOGY

The selection of technology has long been themost controversial element in discussions of MRT.Both costs and performance vary from location tolocation according to stop spacing, vehicle andsystem design, and so on. Table 8.1 gives datafrom systems recently completed or still in con-struction. In broad orders of magnitude, at-gradebusway systems formed by conversion of existingroadway (including vehicles) cost between $1 mil-lion and $8 million per route kilometer, with thecosts increasing to as much as $15 million whereeither the vehicles (as in Quito) or the infrastruc-tures (as in Bogotá) become more sophisticated.Light rapid transit (LRT) costs are typically between$10 million and $30 million, though where the mostsophisticated technology is used in a fully segre-gated system, as in the PUTRA system in KualaLumpur, the costs can approach those of a heavyrail system. Full heavy rail metros cost between$30 million and $100 million, the most expensivebeing fully automatic, fully underground systems.The capacity of the systems varies from busways,which can carry up to 20,000 passengers per hourin the peak direction (pphpd) at an average speed

of 17 to 20 kilometers per hour (km/h), to metros,which can carry up to 80,000 pphpd at an averagespeed of 40 to 50 km/h (where there are long stopspacings). Conversions of existing suburban rail-ways, as in Recife or Linha Sul, Brazil, may offerhigh-potential capacity at reasonable cost,although many are not appropriately located toexploit this potential.

Busways are the least-expensive form of MRT,extensively developed (and now being extended)in Brazil, Colombia, and Ecuador.11 They providesubstantial capacity as trunk carriers in major cor-ridors, as in São Paulo. They can be operated asa high-quality network, as in Curitiba; with elec-tric traction if required, as in the San Mateus-Jabaquara system in São Paulo;12 even retrofittedinto existing road systems, as in Quito.13 Yet, out-side Latin America, they are rare in the develop-ing world. This is partly due to their associationwith what is regarded as a relatively primitivetechnology, with a “bottom-of-the-market”image, not to be adopted when a metro can beafforded or reasonably aspired to. Their capac-ity is seen as limited, due to the small size of vehi-cles and interactions with other traff ic atintersections.14 In the technologically upgradedform of bus guidance systems, they are nowbeing advocated as possible means of efficientlyusing space at bottlenecks, while being relativelyinexpensive to retrofit.15

Many of these perceived problems can be over-come. Well-designed large buses (23-meter,double-articulated buses with five doors are usedin Curitiba), off-vehicle ticketing, passing lanesat bus stops, and even platoon operation canbring effective capacity up to 20,000 pphpd. Theoriginal four-lane busway on Avenida Caracas inBogotá managed 36,000 pphpd, albeit withreduced speed performance.

Good system design and specification of state-of-the-art clean and efficient vehicles, as inCuritiba, can change both the environmentalimage and the reality. Electric trolley vehicles, asused in Quito, can further reduce both air and

MASS RAPID TRANSIT 113

TAB

LE 8

.1

PE

RF

OR

MA

NC

E A

ND

CO

ST

OF

SO

ME

TY

PIC

AL

MA

SS

RA

PID

TR

AN

SIT

SY

ST

EM

S

Bo

go

táC

arac

asB

ang

kok

Mex

ico

Cit

yK

uala

Lum

pur

Tuni

sR

ecif

e Q

uito

(Tra

nsM

ileni

o,

P

ort

o A

leg

reE

xam

ple

(line

4)

(BTS

)(li

ne B

)(P

UTR

A)

(SM

LT)

(Lin

ha s

ul)

Bus

way

pha

se 1

)B

usw

ays

Cat

ego

ryR

ail m

etro

R

ail m

etro

R

ail m

etro

Li

ght

rai

lLi

ght

rai

lSu

bur

ban

rai

l B

usw

ayB

usw

ayB

usw

ayco

nver

sio

nTe

chno

log

yE

lect

ric, s

teel

E

lect

ric, s

teel

E

lect

ric, r

ubb

er

Ele

ctric

, E

lect

ric,

Ele

ctric

, A

C e

lect

ric d

uo-

Art

icul

ated

D

iese

l bus

esra

ilra

iltir

ed

river

less

stee

l rai

lst

eel r

ail

tro

lleyb

usd

iese

l bus

esLe

ngth

(kilo

met

ers)

12.3

23.1

23.7

2929

.7 k

m14

.311

.2 (+

ext

5.0)

4125

Vert

ical

seg

reg

atio

n10

0% t

unne

l10

0%

20%

ele

vate

d10

0%

At

gra

de

95%

at

gra

de

At

gra

de,

par

tial

At

gra

de,

A

t g

rad

e, n

o

elev

ated

55%

at

gra

de

elev

ated

5% e

leva

ted

sig

nal p

riorit

ym

ainl

y si

gna

l prio

rity

25%

tun

nel

seg

reg

ated

Sto

p s

pac

ing

1.

51.

01.

11.

30.

91.

20.

40.

70.

4(k

ilom

eter

s)C

apita

l co

st

1,11

01,

700

970

1,45

043

516

611

0.3

213

(inf o

nly)

25(m

illio

ns o

f do

llars

) o

f whi

ch:

Infr

astr

uctu

re/T

A/

833

670

560

—26

814

920

.032

225

equi

pm

ent

(mill

ions

o

f do

llars

)Ve

hicl

es (m

illio

ns

277

1,03

041

0—

167

1880

(113

veh

s)N

ot

incl

uded

N

ot

incl

uded

o

f do

llars

)(p

rivat

e o

per

atio

n)(p

rivat

e o

per

atio

n)C

apita

l co

st/r

out

e 90

.25

73.5

940

.92

50.0

13.3

11.6

10.3

5.2

1.0

kilo

met

ers

(mill

ions

o

f do

llars

)In

itial

(ulti

mat

e)

20 (3

0)20

(30)

13 (2

6)30

—8

40 (c

onv

oy

160

—ve

hicl

es o

r tr

ains

/o

per

atio

n ho

ur/d

irect

ion

pla

nned

)In

itial

max

imum

21

,600

25,0

0019

,500

10,0

0012

000

9,60

09,

000

20,0

00p

asse

nger

cap

acity

Max

imum

pas

seng

er

32,4

0050

,000

39,3

0030

,000

1200

036

,000

15,0

0035

,000

20,0

00ca

pac

ityA

vera

ge

op

erat

ing

50

4545

5013

/20

3920

20+

(sto

pp

ing

)20

spee

d (k

m/h

)30

+ (e

xpre

ss)

Rev

/op

erat

ing

co

st

—10

020

>10

011

5% in

—

100

100

100

ratio

1998

Ow

ners

hip

Pub

licPr

ivat

e (B

OT)

Pub

licPr

ivat

e (B

OT)

Pub

licPu

blic

Pub

lic (B

OT)

Pu

blic

Pu

blic

un

der

in

fras

truc

ture

,in

fras

truc

ture

,co

nsid

erat

ion

priv

ate

vehi

cles

priv

ate

vehi

cles

Year

co

mp

lete

d

2004

1999

2000

1998

1998

2002

1995

(ext

200

0)20

00 (1

998

pric

es)

Mo

stly

199

0s

— N

ot

avai

lab

le.

No

te: B

OT

= B

uy-o

wn-

tran

sfer

. Tun

is, T

unis

ia; R

ecife

, Bra

zil.

Sour

ces:

Pat

tiso

n 19

99; a

nd B

B&

J C

ons

ult

2000

.

114 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

noise pollution, at the expense of a near dou-bling of total system costs. In cities where roadsare wide (as in much of the former Soviet Union)and the bus industry fairly concentrated, electrictrolley vehicles can provide an affordable andflexible MRT alternative that is acceptable bothto passengers and to the traditional carriers.However, they do require strong political com-mitment and effective public sector planning inovercoming impediments to finance, in ensuringpriority treatment in traffic management and infra-structure design, and in service procurement andsupervision. Systems in Curitiba and Bogotá haveshown that this can be done, but it requires effec-tive forward planning, strong local political lead-e r sh ip , and a degree o f s tab i l i t y andnonpartisanship in policy that is not found every-where.16

Light rapid transit (LRT) ranges from the con-ventional on-street tramways of Eastern Europeand the Arab Republic of Egypt to the sophisti-cated elevated and completely segregated sys-tems of Singapore. LRT is expanding rapidly inindustrialized countries in cities with low corridorvolumes, as a secure and high-quality alternativeto the private car, and sometimes serving as afeeder to heavy rail systems. In developing coun-tries, where the need is for adequate capacityand speed for captive low-income public trans-port passengers, its role is less obvious. Whereit operates at grade, without priority or protec-tion from obstruction by other traffic, it has littleor no performance advantage over busways.17

Although low-cost rehabilitation of on-streettramways may sometimes be worthwhile (anexample is the Bank-assisted rehabilitation inBudapest), they are usually an expensive propo-sition for the volume and speed achieved. Theirmain advantages are that they have less local airpollution impact, signal a more permanent com-mitment to public transport, and have an imagethat triggers support for complementary meas-ures, which buses have great difficulty achieving.

With the exception of the conventional tram sys-tems of Eastern Europe and the former Soviet

Union, LRTs exist, or have been planned, only inrelatively wealthy cities—such as Hong Kong,China; Singapore; Tunis; and Kuala Lumpur—orfor high-income developments—such as the Trende la Costa of Buenos Aires. Some newer road-based light transit systems, such as the Fura Filaguided trolleybus in São Paulo, have lower infra-structure costs than have a similarly segregatedLRT. At the moment, LRTs are usually low-capac-ity prestige systems of doubtful value in poorercities in the earlier stages of development.However, if viewed as mixed systems, partly atgrade and partly elevated or underground, andas interim steps to the creation of a full metro inlarger and more dynamic cities, they may havesome role to play, not least as a means of pre-serving the necessary right-of-way for a subse-quent metro development.

Suburban railways are often well located for radialjourneys. Although even in relatively well-servedcities (Mumbai, Rio de Janeiro, Moscow, BuenosAires, and Johannesburg), they carry fewer than 10percent of trips, they can be very important in pro-viding for longer commuting trips. Existing, butless-well-used lines can be converted to effectivelocal passenger service either in the form of a con-ventional rail service sharing facilities with other railtraffic or by replacement on the same right-of-way by a light rail system.18 Suburban railways canhave significant disadvantages, however.Sometimes, because they were developed beforethe growth of motorization, they have at-gradecrossings that reduce their speed and capacity,and present serious safety hazards. Where thecenter of activity in the city has shifted, the centralstations may not be well located; and in somecases underused rights-of-way become encum-bered by poor people living illegally on publiclyowned land (squatters), which makes redevelop-ment difficult. In some cases the sharing of trackwith freight or long-distance passenger services,as well as frequent grade crossings, also reducestheir capacity.19 Despite these problems, severalsystems could be converted into surface metrosat a fraction of the cost of underground or ele-vated systems, even if there is a need to add an

MASS RAPID TRANSIT 115

underground or elevated link to the central busi-ness district or any other populated area.20 SeveralAsian cities (Mumbai, Delhi, Manila, Bangkok) andAfrican cities (Abidjan, Côte d’Ivoire; Maputo,Mozambique; Cape Town, South Africa) are pos-sible candidates for conversion of suburban rail-ways into modern systems operated with dieselmultiple units or electric multiple units.

The most serious impediments to suburban rail-ways are frequently institutional. When operatedby national rail organizations, suburban railwaystend to be given low priority and are poorly coor-dinated with other urban public transport serv-ices (as in Colombo, Sri Lanka; Moscow; andMumbai). In some cases the weakness of pub-licly owned national rail undertakings leaves theircapacity severely underdeveloped (as in Manila).Recent experience has shown what can beachieved by addressing these problems. A pro-gram of concessioning to the private sector inBuenos Aires has revitalized the system, doublingpatronage over a five-year period while reduc-ing the budget burden of the system by nearly$1 billion per year. In Brazil the transfer of respon-sibility for suburban railways from the highly cen-tralized Companhia Brasileira de Trens Urbanos(CBTU) to local (state) control, together with agovernment-funded rehabilitation program, hasimproved service in most of the major cities,21

and, assisted by a program of concessioning,22

is greatly reducing the fiscal burden (box 8.1).The costs of developing existing rights-of-waycan be very reasonable; there is considerablescope for such developments. Serious attentionshould be given to eliminating the institutionalimpediments to the development of suburbanrailways as part of metropolitan MRT networks.

Metros are usually the most expensive form ofMRT per route kilometer, but have the greatestcapacity and best performance. With 10-car trainsets and two-minute headways, the first HongKong, China, line has carried as many as 80,000pphpd. The Moscow metro has regularlyachieved headways below 90 seconds. SãoPaulo’s East line has consistently carried more

than 60,000 pphpd. However, many metros aredesigned for capacities around 30,000 to 40,000pphpd, and few actually carry more than that.

Costs can vary greatly. Total capital costs can beas low as $8 million per kilometer where an at-grade right-of-way is available for conversion,and can rise to over $150 million per kilometerfor an underground railway in difficult terrain.

A study of recently built metros showed totalcosts per kilometer for the underground systemsin Latin America to be from two to three timeshigher than those for Madrid (table 8.2).23 To alarge degree, this was attributed to differencesin project management arrangements. In Madrid,there was a high political commitment to com-plete the project, full financing was ensured fromthe beginning, and contractors were paid withoutdelay. In most Latin American cases, there weredelays and interruptions caused by lack of avail-able finance and changing political priorities.Moreover, the small but highly experienced proj-ect management team of the Madrid metro com-pany was given full power to make on-the-spottechnical and financial decisions, thereby avoid-ing work delays and ensuring quick progress pay-ments to the contractors. The government wasseen to be a credible client by the contractors,whose prices thus included a lower risk elementthan is sometimes the case elsewhere. Additionalreasons for the lower costs in Madrid have beenascribed to the technology adopted. Civil workscosts were kept low by using the earth pressurebalance method for tunneling, by strong geot-echnical supervision monitoring, and by a stan-dardized station design. Equipment costdifferences were attributed to the choice of con-ventional steel-wheel technology with overheadcurrent collection, and carefully phased acquisi-tion of signal and communications equipment,the technology of which was specified to beslightly lower than that in some of the LatinAmerican cases. Madrid also benefited from theoption of extending a previous supply contract,which permitted the acquisition of new rollingstock at a relatively low cost.

116 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

The financial performance of some of these sys-tems is shown in table 8.3. For a fully privatelyfinanced metro, operating costs only account forabout 40 percent of total cost, with capital chargesaccounting for the remainder. On that basis, onlyHong Kong, China, appears to come close to cov-ering total costs; many do not cover operatingcosts. The table also indicates that, while costsdo differ substantially between systems, the two

main factors affecting financial viability are thecorridor volumes and the revenue per passenger.Although the existence of large external effects(about one-half of the benefits typically accrue toremaining road users) means that economic ratesof return may be positive and acceptable evenwhere the financial return is negative, the samefactors that affect financial viability also affect theconventional measures of economic viability.

BOX 8.1 LEVERAGING URBAN TRANSPORT COORDINATION IN BRAZIL

Responsibility for urban public transport in the Brazilian conurbations has historically been veryfragmented. The federal government owned and operated suburban railways, the states wereresponsible for intermunicipal buses within the conurbation, and municipalities were responsi-ble for intramunicipality bus services. Metros were under either state or municipal control. Policywas not coordinated, especially when the various levels of government were under different polit-ical control. In the early 1990s, as part of a policy of decentralization, the federal governmentdecided to rehabilitate the various urban rail systems and transfer them to the states.

The World Bank financed part of the rehabilitation in several cities (Recife, Belo Horizonte, Riode Janeiro, São Paulo, Salvador, and Fortaleza). As part of the process, the opportunity was takento address the broader issues of urban transport coordination. In addition to funding infrastruc-ture or rolling stock rehabilitation, each project contained the following elements:

• Establishment of a regional transport coordination body• Completion of an integrated urban transport, land-use, and air-quality strategy and a regional

public transport plan• Introduction of financing mechanisms to establish the basis for financially sustainable public

transport• Increased involvement of the private sector in operations and investment in the sector.

The reforms have limitations. The coordinating bodies usually lack executive power, becausemayors and governors are unwilling to delegate their regulatory powers. Even good transportplans do not ensure consistent implementation. The financial provisions may simply be the inclu-sion of a budget line rather than a secure, long-term source of finance—and private participa-tion does not exclude all need for public funding.

Nevertheless, the achievements have already been considerable. In Rio both the metro and thesuburban railways have been successfully concessioned to the private sector. In São Paulo metroand suburban rail systems are being linked, and their development coordinated. In Fortaleza,state and municipality work together to restructure networks and introduce integrated ticketingsystems.

Source: J. Rebelo, based on World Bank internal documents.

MASS RAPID TRANSIT 117

Capital costs, which may account for as much astwo-thirds of total costs, are very largely deter-mined by international prices for the metro tech-nology. In contrast i t is typical for aboutthree-quarters of conventionally measured eco-nomic benefits to be in the form of time savingsfor metro passengers and bus passengers; thesesavings are a function of local incomes. Hence itis not only the financial but also the economicbenefit of metros that depend critically on incomelevels.

OWNERSHIP AND FINANCING

Most metros require subsidy. As table 8.3 shows,some are able to cover operating costs, exclud-ing depreciation of assets; full-cost coveragewould require an operating cost cover ratio ofabout two. While relatively low operating costsoffset low revenues per passenger in the casesof Santiago and Singapore, the differences in thelevel of cost coverage are more strongly influ-enced by variations in revenue per passengerthan in costs. As a result, new metros are likely

to impose a heavy fiscal burden unless the sys-tems are very heavily utilized and fares are rela-tively high. Even a relatively rich city—such asPusan, the Republic of Korea—found its metroto be such a burden that it had to be transferredback to the national government account. It wasestimated that the 70 percent central govern-ment contribution to the proposed Bogotá metro(shelved in 2000) would have required 30 percentof the uncommitted national investment budgetfor the next 10 years. Such a burden might wellpreempt other social expenditures that wouldbe of great benefit to the poor.

In recent years the mobilization of privatefinance has been seen as a way of escaping theadverse fiscal burden, with two systems in KualaLumpur and one each in Bangkok and Manilabeing privately financed. In all cases the proj-ects appear to have been constructed to timeand budget, but all have fallen short of esti-mated passenger demand. The revenue risk inthe Manila project fell to government under thebuild, lease, and transfer (BLT) structureadopted. The two Kuala Lumpur projects have

TABLE 8.2 VARIATION OF COST IN RECENT METRO CONTRACTS

Santiago Costs per kilometera Madrid Caracas Percent line 5 Percent (millions of dollars) extensions line 4 of Madrid extension of Madrid

Total costs per kilometer 31.18 93.56 300 70.1 225

Civil works 19.69 31.41 160 33.68 171

Equipment costs 5.13 22.48 438 11.56 225

Track 1.27 2.67 210 3.57 281

Power supply 0.85 8.45 994 2.36 278

Signaling and so on 0.96 6.51 678 4.36 454

Station equipment 0.61 2.13 349 0.35 57

Escalators and lifts 0.99 2.00 202 0.72 73

Ticketing equipment 0.12 0.68 567 0.20 167

Rolling stock 5.54 25.77 465 17.84 322

Rolling stock (adjusted)b 4.87 10.74 221 8.35 171

Design and management 0.69 4.64 672 8.54 1,237

a. All costs are in U.S. dollars at September 2000 exchange rates.b. Adjusted to capacity of 10,000 passengers per hour, at four standing passengers per square meter. Source: Authors.

118 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

had to be restructured with government effec-tively bailing them out. The viability of theBangkok Transit System (BTS), the first to bepurely privately financed, has not yet beenproven. Furthermore, reliance on private financehas discouraged integration with other modesor other MRT lines, so that the contribution tothe total urban system has been less productivethan it might have been.24

In the case of busways, it has proved difficult tostructure private concessions for the provision ofinfrastructure and services, with failed attempts inSão Paulo (1995) and Bogotá (1996).25 Despitetheir proven operational and financial perform-ance, it appears that without active governmentparticipation in the implementation process, therisk may be too great to attract private finance.26

Bogotá’s recent TransMilenio project thus con-cessioned only the provision and operation of its470 buses, but the physical busway was financedand implemented by the municipal government(box 8.2).

A different approach to private involvement hasbeen employed in the concessioning of existing

metro and suburban rail systems in Buenos Aires,Rio de Janeiro, and other Brazilian cities, and isnow being considered in Mexico City. In BuenosAires the contracts have involved governmentfinancing of major rehabilitation as well as oper-ating subsidies based on winning bids. In theBrazilian cities the rehabilitation took place largelybefore concessioning, under the terms of thetransfer of responsibility for suburban railwaysfrom the federal to state governments. In MexicoCity the concession of 200 kilometers of subur-ban railways is being designed in three gross-cost-type contracts to enable the authorities tocontrol fares within a coordinated system.

Private management has greatly improved serv-ice supply and cost recovery and reduced thefiscal burden. In Buenos Aires between 1993 and1999, metro and suburban rail line coach kilo-meters increased by 75 percent, proportion oftrains on time by 20 percent, and train kilome-ters cancelled fell by 80 percent. Although faresrose by 30 percent in real terms over the period,patronage increased by 125 percent. The resultwas that the subsidy cost per passenger fell by90 percent, from $1 to $0.10. Experience thus

TABLE 8.3 FINANCIAL PERFORMANCE OF SOME METRO SYSTEMS

Operating System Passengers Revenue Cost cost per Revenue length Popu- per per per kilometer to (kilo- lation kilometer passenger passenger (millions operating

City meters) (millons) (millions) (dollars) (dollars) of dollars) cost

Santiago 37.6 4.9 4.92 0.35 0.19 37.8 1.84

Singapore 83.0 4.0 4.67 0.57 0.34 71.9 1.67

Hong Kong, China 82.0 7.1 9.36 0.96 0.61 65.2 1.56

Buenos Aires 47.4 12.6 5.46 0.59 0.43 78.8 1.39

São Paulo 49.2 17.8 9.32 0.62 0.61 65.4 1.02

Seoul 286.9 12.5 6.56 0.38 0.44 64.6 0.87

Pusan 54.2 4.0 4.43 0.39 0.46 103.2 0.83

Mexico City 191.2 18.1 6.66 0.15 0.28 41.9 0.53

Kolkata 16.45 12.9 4.86 0.11 0.23 47.6 0.42

Note: Conversion to U.S. dollars at official exchange rates.Source: Data from annual reports of companies for year 2000.

MASS RAPID TRANSIT 119

BOX 8.2 TRANSMILENIO: BOGOTÁ’S BUS RAPID TRANSIT SYSTEM

As part of a comprehensive urban mobility strategy including promotion of nonmotorized trans-port (NMT) and restriction of automobile use, the municipality of Bogotá, Colombia, has devel-oped a bus rapid transit system called TransMilenio.

The infrastructure of the system includes exclusive busways on central lanes of major arterialroads, roads for feeder buses, stations, and complementary facilities. Trunk line stations are closedfacilities with one to three berths, varying from 40 to 180 meters in length, located in the medianevery 500 meters, on average. Trunk lines are served by articulated diesel buses with 160-pas-senger capacity, while integrated feeder lines are served by diesel buses with capacity of 80 pas-sengers each. To maximize capacity, trunk lines accommodate express services stopping at selectedstations only, as well as local services stopping at all stations. This combination allows the systemto carry up to 45,000 passengers per hour per direction.

Services are operated by private consortia of traditional local transport companies, associatedwith national and international investors procured under competitively tendered concession con-tracts on a gross cost basis. A separate marketing contract covering production and distributionof smart cards, acquisition and installation of turnstiles and validating systems, and passengerinformation and money handling was also competitively tendered. All revenues are deposited ina trust fund, from which operators are paid according to their contracts.

Overall system management is performed by a new public company (TransMilenio S.A.) fundedby 3 percent of the ticket sales. TransMilenio S.A. operates a control center, supervising serviceand passenger access. Each articulated bus is equipped to use global positioning system to reportits location every six seconds. Turnstiles also report passenger movements to the control center,allowing supply to be efficiently adjusted to demand.

The system was developed in less than three years starting January 1988; service commenced inDecember 2000. By May 2001 it carried 360,000 trips per weekday, at a ticket cost of $0.36 and with-out operating subsidies, on 20 kilometers of exclusive lanes, 32 stations, 162 articulated buses, and60 feeder buses. Productivity was high, with 6.21 passengers per kilometer, 1,945 passengers perday per bus, and 325 kilometers per day per bus. Fatalities from traffic accidents involving buseshad been eliminated, some air pollutants reduced by 40 percent, and users’ travel time reducedby 32 percent.

When the first phase of TransMilenio is fully operating, in 2002, more than 800,000 passengersper day are expected on 41 kilometers of exclusive lanes, with 62 stations, 470 articulated buses,and 300 feeder buses. It is intended to expand over a 15-year period to include 22 corridors with388 kilometers of exclusive lanes.

Source: Hidalgo Guerrero 2001.

suggests that private construction and man-agement can yield substantial benefits, but thatthe benefit of private participation is likely tobe greatest within a well-considered and care-

fully planned overall strategy, and where thepublic sector accepts the financial implicationsof such public policy objectives and builds inwell-constructed concession arrangements.

120 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

MRT often raises the value of land near stations,suggesting the possibility of mobilizing financefrom “development gain.” In practice, concen-trated development often fails to occur becauseof the fragmentation of landownership. Evenwhere landownership is concentrated, as in HongKong, China, the direct contribution to the publicfunding of metros has been limited to between10 and 15 percent of the capital costs. Subtleinteractions and incentives may arise where thereis common ownership of the MRT and contigu-ous developments or land, as in some urban rail-ways in Japan. Even then, development may beslow to occur (as in Kuala Lumpur’s PUTRAsystem). Perhaps the clearest lesson of theHopewell saga in Bangkok is that linking trans-port infrastructure development to profits of prop-erty development is a very insecure basis forordered progress (box 8.3).

Where ownership of land contiguous to MRT sys-tems is fragmented, development gain can usu-ally only be recouped through taxation. This hastwo main drawbacks. First, it is only recouped expost and therefore cannot contribute to the ini-tial capital finance. Second, because the defini-tion of both area and magnitude of impact isdifficult, specific betterment taxation tends to be

politically contentious. It is therefore much morecommon for taxes related to MRT-financing tobe system related (such as the French “verse-ment transport” and the Buenos Aires subwaytax) than project related.

The project preparation process must be thor-ough. Premature commitment to unsolicited pro-posals has usually involved some contingentobligations, which have come back to haunt gov-ernment, and a lack of coordination, which hasbeen very damaging to the transport system asa whole. For projected corridor flows exceeding10,000 pphpd, alternative MRT solutions shouldbe carefully evaluated within an integrated urbantransport, land-use, and air-quality strategy. Whenchoosing the technology, the need to maintaina competitive supply environment as the systemexpands should also be borne in mind: unusualproprietary technologies can create serious prob-lems in this respect. A very sober economic eval-uation of the system using cost-benefit analysis(including resettlement analysis, environmentalanalysis, and an analysis on the impact on acci-dents), as well as a multicriteria analysis weigh-ing the objectives to be achieved, should beundertaken. A qualitative analysis of the effectson urban structure should be also prepared. An

BOX 8.3 THE HOPEWELL PROJECT IN BANGKOK

A concession to the Hopewell Company for a 60-kilometer multimodal transport system wasapproved by the Thai government in 1991. It included an elevated toll road, an MRT system, gradeseparation of the Thailand State Railway (SRT) track, local road improvements, and a port rail linefrom Makkasan. The financial attractiveness of the project stemmed substantially from the grantof property development rights on SRT-owned land.

At the time of the collapse of the Bangkok property boom, in 1997, work progress came to astandstill, and the project remained only 14 percent completed for over nine months. The gov-ernment unilaterally announced the termination of the concession in spring 1998, an actionrenounced by an official panel in March 1999. Negotiations for continuation of the project havebeen based on the elimination of the toll road, given the completion in the interim of other roadschemes that reduce its attractiveness.

Source: World Bank 1999.

MASS RAPID TRANSIT 121

analysis of the financial impacts and breakevenfares will always be required, and should includeanalysis of the impact of the breakeven fares onthe poor along with their willingness to pay.Finally, the impact of the project on state financesmust be evaluated.

PUBLIC TRANSPORT INTEGRATION

The planning and evaluation of many mass tran-sit projects presume that there will be effectivemodal integration, including the creation of appro-priate interchange facilities (which the Bankfunded in the case of Pusan metro and is pro-posing to fund for Manila LRT3) and bus servicerestructuring. If the objective of an MRT invest-ment is primarily to improve road transport con-ditions, the restructuring of public road passengertransport may be critical to the achievement ofthe desired objective.

There are two main problems in this respect. First,it may take many years for the posited restruc-turing to take place. There are a number of rea-sons for this.

• Fragmentation of operational responsibilitybetween modes (and in the case of bus sys-tems, within the mode) means that there isno initial institutional responsibility for pro-viding appropriate interchange.

• Jurisdictional fragmentation often accentu-ates this, with responsibility for rail-basedmodes resting with state or central govern-ments but responsibility for buses restingwith the municipality.

• Bus services are often regulated by agen-cies that operate largely independently ofthe other transport institutions.

• Bus operators, who are often a powerfulpolitical lobby, may resist the reduction intheir overall market share.

• Restructuring of the formal bus services, ifnot clearly seen as beneficial by the vastmajority of passengers, may be underminedby the emergence of informal paratransit

services retaining direct point-to-point con-nections on a “many-to-many” (from manyorigins to many destinations) basis.

A second main problem is that, even if the insti-tutional arrangements are conducive to an organ-ized physical restructuring, it may not be seen aspolitically or socially feasible because of its effectson some disadvantaged groups. Restructuringof bus services to feed rail stations increases thenumber of interchanges, imposing both inter-change time penalties and fare penalties wherethere is no integrated fare system. In order toovercome this impediment, attempts to developphysically integrated systems in major Braziliancities have been accompanied by integration offares and careful attention to fare structures toavoid adverse distributional consequences (seesection on pricing below). The key to effectivemodal integration is the existence of a strongregional coordination authority backed by thedifferent levels of government.

PRICING

Public transport serves very disparate markets.In many large cities, it serves the basic movementneeds of those without private transport. In morecongested cities, it may also aim to attract com-muting trips of higher-income car owners orpotential car owners. The problem is that the twomarkets are likely to require quite different priceand quality combinations. For buses this can bereconciled by the provision of higher-quality serv-ices (air conditioned, seat only, limited stop) onthe same infrastructure as that for more basicservices at lower costs. Despite the possibility ofhaving two classes of service on the same train,it is more difficult to price-discriminate effectivelyfor rail-commuting journeys.

There are various ways of attempting to addressthis disparity, including flat fares to allow theperipherally located poor to have reasonableaccess to centrally located employment; person-related subsidies, such as the “vale-transporte”

122 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

(subsidy provided through employers to employ-ees) in Brazil; and integrated fare systems to elim-inate the disadvantage of modal interchange.These devices are discussed in chapter 10. Theimportant point to bear in mind with respect toMRT strategy is the high level of fixed costs ofthe rail systems, and the substantial externaleffects and interactions between modes. It istherefore not advisable to take a purist view thatall modes should be independently self-financing.Cross-modal financial transfers may certainly bejustifiable in such circumstances. It is often truethat motorized road users do not meet the fullcosts of the infrastructure that they use. Insistenceon “pure” private financing of public transportinfrastructure, as in the BTS system in Bangkok,may result in the adoption of price levels andstructures that maximize revenues at low trafficvolumes, hence lose substantial external bene-fits and exclude the poor from use of the system.It is equally important, however, to avoid arrange-ments that effectively tax bus users to subsidizea minority of rail users.

The fact that some subsidies may be efficient andacceptable does not mean that any subsidy isefficient and acceptable. The minimum criteriafor subsidy should include the following:

• The subsidy should not be open-ended butshould be embodied in a contract.

• The right to provide subsidized serviceshould be subject to competitive tender.

• The level of acceptable subsidy for a serv-ice or agency should be subject to explicitcost-benefit appraisal.

• The cost of the subsidy must be fiscally sus-tainable.

CONCLUSIONS: TOWARD A STRATEGY FOR MASS RAPID TRANSIT

MRT plays an important part in maintaining theviability and environmental quality of very largecities, but it can also be expensive and canimpose a severe burden on municipal finances. It

is therefore important that a wide range of dif-ferent technologies be considered in the MRTfamily, and that cities adopt technologies appro-priate to both their physical and their financialsituations. The main elements of a strategy forMRT would therefore seem to be as follows:

On planning • Several MRT technologies should always be

considered and the selection among thembased on a thorough and systematic com-parison of their costs and benefits, and theirfinancial sustainability.

• The interaction of MRT with land use, andits sheer financial magnitude, requires itscareful integration into the planning of met-ropolitan structure, transport, and financewithin a comprehensive long-term structureplan for the city.

• The public sector must set strategy, identifyinfrastructure projects in some detail (includ-ing horizontal and vertical alignment, andstation locations) and confirm the accept-ability of environmental consequences, tar-iffs, and any contingent changes to theexisting transport system.

• The public sector must acquire the neces-sary land and rights-of-way, ensure develop-ment permissions, commit funding, andprovide some necessary guarantees. If devel-opment at terminals is desired, it may needto facilitate consolidation of landholdings.

On finance• There must be a comprehensive financial

plan within which the costs of infrastructureand publicly funded operations are foreseenand securely provided for.

• MRT systems should normally be incorpo-rated in citywide price-level and structureplans.

• Especially when private finance is involved,MRT investments should be consistent withan approved city structure plan; oppor-tunistic development on an ad hoc basis hasusually proved to be damaging to welfare,and ultimately costly to the budget.

MASS RAPID TRANSIT 123

• Financing for project implementation mustbe fully ensured to avoid the delays and costoverruns that have plagued past schemes.

• The full cost of new mass transit investmentson municipal budgets, on fares, and on thepoor should be estimated in advance; devel-opment should only proceed on the basisof a sound financial plan and a reasonablecommitment from funding agencies for anyplanned financial support.

• No commitment to major expenditure on newsystems should be made in the absence ofsecure funding to complete the investment.

On management and pricing • Modes of public transport require physical

coordination (to achieve convenient modalinterchange) and fares coordination (to keeppublic transport attractive and to protect thepoor). Stakeholders need to agree on acomprehensive transport strategy plan,within which the relationship between MRTand other modes (both physical and finan-cial) are understood.

• There must be strong political support andcompetent implementation management,with arrangements put in place to facilitatecoordination between multiple public agen-cies.

• Modally integrated fare schemes should beassessed for their impacts on the poor.

NOTES

1. This is particularly likely to occur when thosemaking the forecasts have a positive incentive tobe optimistic—for example, where free counter-part funding is available from central governmentfor “promising” projects, or where private devel-opers want to present a positive view to prospec-tive lenders.

2. That approach is discussed in a recent WorldBank discussion paper (Mitric 1997).

3. Fouracre, Allport, and Thompson 1990;Gardner, Cornwell, and Cracknell 1991; andGardner 1993.

4. Halcrow Fox 2000a.

5. Rebelo 1996.6. PADECO 2000.7. For example, the “9th of July Busway” in São

Paulo, despite being a cost-efficient mass people-mover, had negative environmental and commu-nity impacts, including accidents to pedestriansentering or leaving the system. While these effectscan be minimized by changes in fuel, more pedes-trian overpasses, and maybe grade separation,these all increase costs.

8. Note, however, that there are increasingefforts to reduce air pollution through innovativebus propulsion systems, including natural gas,clean diesel, hybrid diesel and electric, and fuelcells (see chapter 4). Moreover, two recentlydeveloped busways, in Quito and São Paulo areoperated by electric trolleybuses.

9. The elevated BERTS (Bangkok ElevatedRoad and Train System) in Bangkok has given riseto sufficient concern about the “tunnel effect” ithas had on trapping the road-generated con-gestion below it that the government of Thailandhas resolved that all future rail systems in the cen-tral area of the city be underground.10. Fouracre, Allport, and Thompson 1990.11. Total costs for a 32-kilometer system ofbusways in Bogotá, including 470 articulated trunkbuses and 1,000 feeder buses to operate off thebusway, were recently estimated as approximately$8 million per route kilometer (Ardila and Menckhoffforthcoming).12. Rebelo and Machado 2000.13. The critical feature when a phased devel-opment of technology is planned is that the ini-tial land reservation and the subbase constructionmust be sufficiently strong to support subsequentexpansion.14. Congestion was experienced at stops onAvenida Caracas in Bogotá despite there beingtwo lanes in each direction with 400 small-medium buses per hour; the newly developedTransMilenio system will improve performanceby the exclusion of small vehicles.15. Quinn 1998.16. For more details, see Halcrow Fox 2000a.17. The highest capacity for a true street-running system is the 12,000 pphpd achieved on

124 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

the Alexandria-Madina line with 275 passengervehicles operating every 80 seconds, albeit at avery low speed of 6 km/h. The highly segregatedTunis system has a capacity of 13,000, and the 95percent segregated Alexandria-Rami line has acapacity of 18,000. All currently have patronagemuch below these theoretical capacities. Thelargely segregated Tuen Mun system in HongKong, China, is reported to have a capacity ofover 25,000.18. Caetano Roca Giner 2000.19. An important component of a WorldBank–funded urban transport project in BuenosAires was a program of grade separation of rail-road crossings in the city.20. Such a conversion is being done, at relativelylow cost, in Fortaleza, Brazil.21. Rebelo 1999b.

22. Rebelo 1999a.23. BB&J Consult 2000.24. For example, physical transfer between dif-ferent rail lines is very difficult in Kuala Lumpurand Manila, and between rail and bus in Bangkok,Kuala Lumpur, and Manila. Moreover, there is nofare coordination in any of these cases.25. The city of São Paulo awarded concessionsfor a 241-kilometer bus network, but none ofthe winning bidders could obtain financing toimplement the schemes. In 1995 the city did suc-cessfully concession the São Mateus-Jabaquaratrolley busway, which benefited from the factthat most of the infrastructure was already inplace. In Bogotá in 1996, the Metrobus consor-tium failed to close the financing necessary toimplement.26. Menckhoff and Zegras 1999.

THE IMPORTANCE OFNONMOTORIZED TRANSPORT

Nonmotorized transport (NMT) has an unam-biguously benign environmental impact. In manycities it is the main mode of transport for thepoor, and in some a significant source of incomefor them. It therefore has a very significantpoverty impact. Where NMT is the main trans-port mode for the work journeys of the poor, it isalso critical for the economic functioning of thecity. Despite these obvious merits, NMT hastended to be ignored by policymakers in theformulation of infrastructure policy and positivelydiscouraged as a service provider. The purposeof this chapter is to understand why that hashappened, and in light of the evidence on itscharacteristics, role, costs, and benefits, to sug-gest a framework within which NMT’s potentialmay be better exploited.

Some governments appear to have an ideo-logical preference for motorized over NMTbecause they regard it as technologically moreadvanced. The World Bank concern for NMT iscertainly not based on any ideological prefer-ence for low technology in urban transport.Indeed, the environmental impacts of motor-ized transport are often accentuated by out-dated engine and fuel technology, while modern

nonmotorized vehicles may benefit fromadvanced materials technology. Rather anexplicit strategy for NMT is necessary to redressa historic vicious policy circle that has biasedurban transport policy unduly in favor of sacri-ficing the interests of pedestrians and cycliststo those of motor vehicle users. Because of thispolicy, NMT becomes less safe, less convenient,and less attractive, making the forecast declineof NMT a self-fulfilling prophecy. That process isunacceptable, because it stems from a failureto recognize some of the external effects ofmotorized transport that distort individual choiceagainst NMT, and hence militates particularlyagainst the poor who do not have the means touse even motorized public transport.

The two major modes of NMT are walking andvarious forms of cycling, which can be personal orpublic transport. The NMT public transport sector,which is particularly important in South Asia, com-prises many load-carrying variants. Particularly inhigher-income countries, many people also walkor bicycle for exercise and pleasure. We excludesuch activities from our consideration here toconcentrate on NMT only as a mode of neces-sary transport. NMT accounts for between 40 and60 percent of all trips in several major cities inAsia. In the poorer cities in Africa, that propor-tion is even higher.

THE ROLE OF NONMOTORIZED TRANSPORT 125

THE ROLE OF NONMOTORIZED TRANSPORT9Despite its economic importance to the poor—both as a mode of transport and a

source of income—and its environmental advantages, the potential of

nonmotorized transport is often unmobilized or even actively suppressed. A

combination of infrastructure investment, traffic management, and financial

measures can make nonmotorized transport safer and more attractive, to the

benefit not only of the very poor, who are economically captive to nonmotorized

transport, but also of the less poor.

126 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

WALKING IS TRANSPORT

For very short trips, walking is the main mode oftransport in most societies, rich or poor. Indeed,most trips in all countries involve some walking asaccess and egress to the main mode. The modalshare of walking can be very high. Recent studiesshow that between 25 and 50 percent of trips inthe major Indian cities,1 and around 50 percent ofall trips in major African cities, are entirely on foot,and that trips undertaken primarily by publictransport also involve significant walking dis-tances. In medium and smaller cities, the shareof all-walking trips increases to 60 to 70 percent.Clearly, walking dominates for shorter trips, buteven in terms of distances traveled, walkingaccounts for over 50 percent of all trips inMorogoro, Tanzania.2

The political attitude toward pedestrians is oftenneglectful or curiously hostile. Pedestrian spaceis continually being eroded. Fewer than one-halfof the major roads in most Indian cities have side-walks, and those that exist are frequently occu-pied by street vendors, encroached upon byshop premises, or blocked by parked cars,motorcycles, and bicycles.3 As city authoritieshave found it difficult to manage and controlstreet market and footway activities, the trendhas been toward getting rid of them altogether,rather than taking a functional approach to roadhierarchy, whereby the functions of some roadscould be for pedestrians and market activitiesand not for fast-flowing motor vehicle traffic.Whereas in many industrialized countries theadvantage of pedestrianization of shoppingstreets has been recognized by shoppers andtraders alike, it remains difficult to give priority topedestrians in developing countries.

In some very rapidly growing countries, such asChina, there has been an attempt to plan forpedestrians—though planning for pedestrianshas lagged behind planning for vehicles.Attempting to cater to fast and free-flowingmotor vehicle traffic has resulted in the gradualphysical segregation of pedestrians by over-

passes, subways, and barriers, many of which arepoorly designed. Locations are often chosen forconvenience of construction rather than to bestsatisfy pedestrian desire lines. While physical seg-regation can provide safe facilities for pedestri-ans to cross roads, it also results in severance andinconvenience for them.

WALKING ON WHEELS

Bicycles are a desirable mode of transport inmany cities, but it is important to analyze who isusing them, what the prevailing social and polit-ical attitudes are to the use of the mode, andwhether there are particular obstacles to theiruse by women.

WHO BICYCLES? The bicycle represents a trade-up from walking. Inmany of the poorer developing countries, onlymiddle-income households can afford bicycles.For example, in 1996, 54 percent of bicycle usersin Delhi had a monthly family income of over 2,000rupees (about US$41), and only 19 percent of theusers earned less than 1,500 rupees (about US$31).A similar survey in León, Nicaragua, showed 89percent of users to be relatively well-off.

Data from Guangzhou, Delhi, León, and Accrashow that the NMT share of each age cohort cor-relates closely with its share in total mobility, theage cohort 25 to 35 dominating both NMT useand overall mobility. In Vietnam, however, nearlyall secondary-level schoolchildren go to schoolby bicycle, while motorcycles are rapidly takingthe place of the bicycle as a means of transportfor the age group 25 to 35. Men dominate theuse of bicycles in most countries, with the excep-tion of China and Vietnam.4

In some large countries, the proportion of bicy-cle trips has declined, and is continuing todecline, as incomes increase and as the perceivedsafety and security of cycling diminishes. Even intraditional cycling cities, such as Guangzhou andDelhi, those using NMT are apparently doing so

THE ROLE OF NONMOTORIZED TRANSPORT 127

because there is no affordable alternative—thatis, they may be viewed as (at least temporarily)captive passengers. But there are exceptions. Insome European countries, already at high incomelevels, the use of bicycles as a mode of transportis steady or increasing, both as a main mode andas a subsidiary or feeder mode.

ATTITUDES TOWARD THE BICYCLE Some insight into attitudes toward cycling canbe obtained from a survey carried out in five dif-ferent cities across the world (Accra, Delhi,Gangzhou, León, and Lima) in 1996.5 Most bicy-cle users in these cities preferred bicycle to busprimarily because it was less expensive, but themajority also found it more flexible in routing,faster, and more reliable. The survey did suggest,however, that many would change to motorcy-cles except for the cost. In Ouagadougou, a cityalmost exclusively dependent on two-wheeledtransport, and having made better provision forit than most, bicycles were still clearly viewed asan inferior mode to be abandoned as soon asthe household could afford a motorcycle.

In two of the cities, public transport users werealso asked about the relative merits of bicycleand public transport. Interestingly, most of theperceived benefits of cycling were the same asthose perceived by the cyclists, namely, cost,speed, and flexibility. The main disadvantagesthat this group perceived (which by definitionwere clinching in their decision to use publictransport) were the danger of cycling, the risk ofthefts and assaults, the lack of bicycle paths, andthe motorists’ lack of respect for traffic laws (alsoby implication safety concerns). Only a minorityquoted excessive trip distance or topography asclinching factors. One implication is that if thesefactors could be overcome, the use of NMT mightbe even greater. Another, less-comfortable impli-cation is that if private motorized transport couldbe afforded and were to have the same flexibil-ity as NMT, it would be preferred. Certainly thatwould seem to be consistent with the recent rapidincreases in motorcycling in the richer develop-ing countries.

The attitudes of public authorities toward themodes are often inconsistent with individuals’responses. While efforts have been made to seg-regate motorized and NMT in China, relativelylittle is being done elsewhere on a national basis,and improvement in the conditions for NMTdepend heavily on the enthusiasm of specificmunicipal governments. Only a few (as for exam-ple, Bogotá under Mayor Penalosa) haveattempted major initiatives. As far as public NMTis concerned, a number of governments—mostnotably that of Indonesia—have taken positiveactions to eliminate it.

The reasons for these official attitudes appear tobe rather complex, as Chinese experience shows(box 9.1). There is almost certainly a lack of inter-est by engineers (a lack of interest shared, untilcomparatively recently, in industrialized coun-tries), who favor dealing with more technicallyrewarding road and bridge design. Police oftenfocus on the difficulties of enforcing NMT routesand the lack of respect of cyclists for traffic reg-ulations. Similarly, the richer and more politicallyinfluential classes are likely to be car users andto have a vested interest in reducing the nuisanceoffered by slow-moving and congesting NMT.Lack of transport-planning skills and design solu-tions appropriate for the large NMT volumesfound in developing countries also contributes.Where local government is dependent on cen-tral government for road infrastructure funding(not the case in China), this may also limit localinitiatives in providing for NMT.

WOMEN AND BICYCLINGIn many countries women are largely excluded fromthe use of bicycles. This is partly a consequence ofgeneral economic and social norms, which can beaddressed by community awareness programs toovercome cultural constraints to women’s use ofNMT. There are often more direct transport policyconstraints—associated with access to credit mech-anisms, design of vehicles, and personal safety andsecurity during travel—that may impinge more onwomen than on men and that should be addressedin a policy package.

128 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

THE POLICY PACKAGE

In most countries NMT has developed sponta-neously and remains largely outside the normalprocesses of transport planning. Provision for NMT,if made at all, tends to be “retrofitted” to existinginfrastructure, and to concentrate on minimizingthe disturbance that it causes to the flow of motor-ized traffic. The consequence is that the provisionsare not only expensive but often inconvenient forNMT (for example, open pedestrian footbridgesfor crossing busy roads). Even where it is neces-sary to retrofit, the chances of designing an attrac-tive NMT package are much greater when there isa willingness to modify elements of the system notoriginally designed for NMT. For example, meas-

ures of traffic calming to reduce speed differen-tials between motorized and nonmotorized traf-fic may be an essential requirement if pedestriansare to be kept safe and bicycle traffic is to beattracted to unsegregated roads.

The more successful schemes, whether provid-ing for NMT as a feeder mode or a main mode,have been incorporated in initial urban systemdesign. For example, the plan for Tama NewTown in Tokyo provides for pedestrian and bicy-cle access to town centers and railway stationscompletely segregated from vehicular road traf-fic. The basis for such comprehensive attention toNMT is a combination of national strategy andlocal implementation planning.

BOX 9.1 SOME RECENT EXPERIENCE WITH FACILITIES FOR CYCLING IN CHINA

In Shanghai, the development of NMT routes has been slow; the pace of development has super-seded many proposals. But such problems did not affect the rapid development of motorizedvehicle routes. In retrospect, it would appear that the Shanghai agencies saw the creation of NMTroutes as a way of increasing the capacity and facilitating the operation of motorized vehicleroutes, rather than providing safer and easier passage for NMT on a network of routes. The NMTroutes suffered from problems of motorized vehicle access and parking.

In Guangzhou, the piecemeal development of elevated city-center routes has severed NMT andpedestrian routes, and again, the development of segregated NMT routes was seen by theGuangzhou agencies as a way of improving capacity for motorized vehicles.

In Urumqi, NMT volumes have decreased dramatically since 1992 with the introduction of minibuses,particularly in the central area. Weather and geography also result in NMT volumes lower thanin other Chinese cities. Consequently, existing and previously physically segregated NMT lanesin the city have been converted into service roads and footways. Cycle lanes on central area roadsare being converted to bus lanes.

In Beijing, NMTs are increasingly being squeezed by motorized vehicle parking in physically seg-regated NMT lanes and by the reallocation of space in wide NMT lanes to through motorizedvehicle traffic. On the Second Ring Road, the outside half of the NMT lane has been reassignedto motorized vehicles, and the inside half is used by buses and taxis. NMT parking at work unitsis increasingly being moved to distant, inconvenient locations to provide more convenient spacefor motorized vehicle parking. A recent high-profile closure to NMTs of a commercial street atXidan in Beijing epitomizes current practice.

Source: Frame 1999.

THE ROLE OF NONMOTORIZED TRANSPORT 129

Unless bicycle infrastructure networks are suffi-ciently dense, continuous, and direct, and bothlinks and intersections perceived as safe, theimpediments to cycling will continue to be strong.That will involve understanding desire line flowsand identifying missing links, safety black spots,and other major impediments to bicycle use.Many attempts to mobilize the potential of NMThave failed because the attempts did not con-tain the minimum package of elements to ensurethat NMT became an attractive proposition to asignificant number of people. Given the need fora safe, secure, and direct transit, this minimumpackage would need to include the provision ofsafe, adequately segregated infrastructure; directroutings without major intersection conflicts withmotorized traffic; secure bicycle parking to pre-clude theft; and financially affordable means ofvehicle procurement. Hence, it is desirable thatthere be a local cycling master plan that shouldbe the basis both for planning infrastructurespecifically for bicycles and for incorporatingcycling into general traffic infrastructure and man-agement planning. Such plans have been drawnup and implemented in Dutch cities such asDelft.6 An ambitious bicycle master plan forBogotá, Colombia, has recently been published(box 9.2).

Local planning alone may not be enough. At thenational level, it is necessary to identify objec-tives for the sector and to ensure that the crucialelements of the facilitating framework that canonly be handled at the national level (such as thecreation of a legal basis for traffic management,promotional campaigns, financing instruments,and so on) are put in place. The Dutch BicycleMaster Plan specifies a general objective andidentifies a number of “spearheads” for action;the plan is an element of the long-term transportstrategy for the Netherlands. A similar centralgovernment initiative has recently been taken bythe government of South Africa in establishing anational bicycle transport partnership (box 9.3).

INFRASTRUCTURE POLICIES Network planning must be supplemented bydetailed road section planning (what to do withthe layout: separation or mixing) and junctionand intersection planning. This will involve a com-bination of measures to determine the function,physical characteristics, and use of each elementin the network.

The starting point is function. There are severaldifferent functions (such as access, distribution,or transit) for each kind of mode. Since the inter-

BOX 9.2 THE BOGOTÁ BICYCLE MASTER PLAN

In year 2000 the municipality of Bogotá, Colombia, published a master plan for bicycles in thecity. The plan includes the construction of 320 kilometers of cycleways over a nine-year period atan estimated cost of $120 million. It also provides for the necessary ancillary infrastructure, includ-ing bicycle parking, urban street furniture, landscaping, and traffic signals. The plan has beensubject to detailed economic and environmental appraisals. For a shift in modal split to the bicy-cle of 2.5 percent, an economic rate of return of 15 percent was estimated. Higher shifts give cor-respondingly higher returns.

The launch of the first phase of the plan, for the construction of 200 kilometers of cycle track at acost of $50 million, was accompanied by an ambitious marketing effort. This showed the programto be well integrated with parallel development of the TransMilenio urban public transport schemeand other transport facilities, as well as offered links to neighboring municipalities.

Source: Mauricio Cuellar, from local Bogotá press reports .

130 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

ests of various modes may differ (and conflictsare endemic), the planner must develop a func-tional classification of roads and road hierarchyand identify the appropriate measure to recon-cile conflicts. If the functions for different modesare incompatible—such as the use of “rat-runs”through residential areas as segments of maintrunk movement routes—it will be necessary tojudge on priorities and modify both physicalcharacteristics and regulation of use accord-ingly. Some cases (for example, major arterialsdesignated for faster, longer-distance move-

ments or urban busways) might justify the exclu-sion of nonmotorized vehicles on both efficiencyand safety grounds. Nevertheless, even in thosecircumstances, it is important that steps aretaken to avoid serious severance of short-dis-tance movements. Moreover, any decisions torestructure roads (for example, to introduce newrestraints on categories of use or to take awayprotected bicycle lanes) should be appraised interms of the net benefit to all types of user andnot merely in terms of the speed of motorizedtraffic.

BOX 9.3 SHOVA LULA (RIDE EASY): A NATIONAL BICYCLE TRANSPORT PARTNERSHIP

In his 2000/01-year budget speech, the minister of transport of South Africa announced a plan topromote greater self-reliance through a national bicycle transport partnership. The minister him-self is the champion of this partnership and has the overall authority to define its course. Thenational department of transport is responsible for overall project management and implemen-tation.

At the core of the partnership is a project called the National Bicycle Transport DemonstrationProgram. The aim of this program is to demonstrate the transport benefits of bicycle promotiontargeted at low-income users in relatively low-traffic rural and medium-size city areas. It consistsof a road show to raise awareness and to leverage support, including targeting women, girls, andyouths; procuring and providing low-cost new and used bicycles; training in riding and maintenanceskills; developing a container microbusiness outlet to support users; reviewing infrastructuresafety; and providing detailed training, planning, and evaluation exercises.

This program is funded by the national department of transportation (NDoT), with the aim ofleveraging additional support from other parties. The objective of Phase 1 (2000/01) is to imple-ment at least 10,000 bicycle transport packages at one or more sites in each province across SouthAfrica. The objective of Phase 2 (2001/02) is to implement a further 15,000 bicycle packages atthe nine existing and eight new sites. Phase 3 (2002/03) aims to implement 50,000 bicycle pack-ages at the 17 existing and 18 new sites.

In addition to the national demonstration program, the partnership has also developed rural andurban pilot projects. The urban project is a partnership between the NDoT and the Midrand LocalCouncil. As part of their Ecocity initiative, Midrand has committed 300,000 rands (about US$30,675)to promoting 1,000 subsidized bicycle transport packages in the township of Ivory Park, east ofJohannesburg, together with a container microbusiness. Midrand, together with its partners, hasalso designed six kilometers of bicycle infrastructure in Ivory Park in order to ensure safety and max-imize the promotion of township cycling.

Source: De Langen and Tembele 2000.

THE ROLE OF NONMOTORIZED TRANSPORT 131

Intersection design and design of motor vehicleaccess to buildings on NMT routes are bothextremely important, particularly, as in China,where bicycle volumes may be very high.

It may often be possible to reconcile functionsof a road section by changing its physical layout.Motorized and nonmotorized traffic can coexistin three ways.

a. Full integration gives no exclusive right orspecial protection for bicyclists or pedestri-ans using a mixed road and relies on drivingbehavior to protect the more vulnerable cat-egories.

b. Partial segregation reserves a strip on thecarriageway for bicyclists or pedestrians, butdoes not protect it physically.

c. Full segregation gives exclusive rights topedestrians or cyclists and makes it physi-cally difficult for motorized traffic to trespasson that right. Even in this case, however, itis likely that there will be at best only partialsegregation at major intersections.

These alternatives are in ascending order of cap-ital cost but descending order of enforcementcost. The choice between them will depend onthe traffic mix, primary road function, relative freeflow speeds, and traffic volumes. Experience inAfrica (Ghana) and Latin America (Peru) haveshown that if the volume of bicyclists is too low,it will invite infringement, at least by pedestrians,particularly if there is no pedestrian alternative.

Where NMT volumes are high, it will not only besafest but also most efficient in terms of travelspeed for all categories of traffic to provide forfull segregation. It may then also be efficient, asincreasingly practiced in larger cities in China, toprovide for grade separation of traffic at inter-sections, though there may be a danger that, forcost reasons, cyclists are assigned a cumbersomeroute that reduces the relative comfort of thatmode. At the very least, it will be justified to pro-

vide for separate NMT phases in signal-controlledintersections.

Where motorized traffic volumes are low and thelocation sensitive (for example, in residentialareas), it may be more appropriate to use traf-fic-calming devices to slow all movements tospeeds at which they are safely compatible (asin the Dutch “woonerf”—sharing space betweenvehicles and pedestrians). The adoption at thenational level of advisory standards for trafficseparation is one basis on which a more NMT-sensitive attitude to road design can be effec-tively disseminated.7

TRAFFIC MANAGEMENTSurvey evidence makes it clear that safety is themost critical feature for the preservation anddevelopment of NMT. Physical separation ofspace for NMT can be made largely self-enforc-ing. Where there are no physical barriers toinfringement, separation will only work if thereis a sufficiently powerful combination of moni-toring, enforcement, and legal penalties, par-ticularly on vehicle infringements to run or parkin NMT space. Most difficult of all is the protec-tion of NMT in shared space; evidence in manycountries suggests that this will have to be donethrough strong, physically enforced, traffic calm-ing. Experience in the Netherlands and else-where shows that traffic-calming measures inneighborhoods and shopping or market areasare among the most effective supports to NMTprograms.

Theft is a strong deterrent to bicycle ownership.A household survey in Guangzhou in 1995revealed that each household had on averageone bicycle a year stolen from it, and that 62 per-cent of these thefts occurred in residential areas.Protection against theft has proved elusive. Nearlyall the stolen bicycles in Guangzhou were locked.They were also registered with the police, whowere, however, unwilling even to record the theftsand who recovered fewer than 10 percent of thestolen bicycles. (In Jiangmen City, where a newbicycle number plate practice is in place, results

132 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

are different.) Guarded parking facilities, commonin Japan and the Netherlands, offer better pro-tection.8 Unfortunately, most cities do not havesuitably located or sufficiently inexpensive park-ing facilities at major destinations (stations, mar-kets, offices, and so on).

SAFETY EDUCATION AND TRAININGAn important requirement of a strategy to pro-mote NMT is to establish and make known to thegeneral public the rights and responsibilities ofpedestrians and cyclists, as well as the traffic rulesthat deal with safety. In many Organisation forEconomic Co-operation and Development(OECD) countries, these rights and rules are clear.In contrast, in many developing countries cyclistsand pedestrians are often mentioned in legisla-tion only for definition purposes and have nospecified legal rights in the use of road space.Traffic law should clearly define the rights andresponsibilities of nonmotorized users of roads,as well as those of motorized users.

Establishing rules is only one part of the solutionto the problem. The second part is to get themknown and applied. Training and educating chil-dren to understand more about traffic rules canbe addressed both by incorporation of such train-ing in the school curriculum and by campaignsoutside school. In the longer term, knowledgeof the rights of pedestrians and cyclists shouldbe incorporated in driver license testing. In theshorter term, the problems are twofold: manyexisting drivers are unlicensed, and few policehave been educated to think of the safety ofpedestrians and cyclists as a responsibility of driv-ers of motorized vehicles. Treatment of NMTshould thus be a central part of comprehensiveroad-safety programs (see chapter 5).

MUNICIPAL INSTITUTIONS ANDORGANIZATION

Provision for NMT in urban areas is almost exclu-sively a municipal responsibility. As such it suf-fers from the typical paucity of municipal funding

in developing countries. The virtual impossibil-ity of charging users directly for the infrastruc-ture provided accentuates this disadvantage.Even secure bicycle parking, which is providedcommercially in some industrialized countries, isdifficult to finance in developing countries.

In some of the higher-income developing coun-tries, it may be possible to learn from more indus-trialized countries, where adequate provision forNMT—especially walking—is a planning require-ment for commercial shopping mall development.But that requires both a degree of effectivenessof development controls and a political commit-ment to adequate provision for NMT. The criticalquestion is how such commitment and capabil-ity can be developed.

One approach, being pursued in some Africancities, is to set up a special-purpose multidisci-plinary and interdepartmental team within themunicipal authority to take initiatives, to plan,and to implement interventions. Such a teammight support intensive user participation in theplanning and design process and focus on widelysupported interventions with a high benefit-to-cost ratio (value for money). A template for suchan approach has been provided for African cities.9

That approach is most likely to be effectivewhere the modal share of NMT is large and itsimportance is already recognized. In less pro-pitious circumstances, it may be more impor-tant to create alliances rather than to risk furtherfragmentation of responsibility for urban infra-structure. It may then be more effective to buildcapacity by embedding NMT expertise in exist-ing forums, such as municipal engineeringdepartments, national road-safety councils, roadfund administrations, and so on. In either case,the starting point needs to be recognition ofthe potential for NMT and a commitment to pro-viding for it.

STAKEHOLDER PARTICIPATION It might be expected that user and stakeholderparticipation would improve the quality of the

THE ROLE OF NONMOTORIZED TRANSPORT 133

decisions by exploiting local knowledge anddesires, and spotting and solving potential prob-lems, as well as by smoothing the execution ofpublic sector interventions. In practice, experi-ence with user participation in NMT schemes hasbeen very variable. In the NMT project in Lima,staff and authorities showed little commitmentto participation, and ad hoc meetings wereorganized only in response to local protests.Similarly, the Kenyan authorities lacked commit-ment in implementing the Sub-Saharan AfricaTransport Program (SSATP), and success was lim-ited. However, in Tanzania the SSATP results werepositive, partly because of the urgency of theproblems and the expectation that the outcomeof the entire process would be positive for theuser group involved (box 9.4).

FINANCINGMany of the impediments to cycling and walkingarise from the inadequacy of infrastructure thatis not addressed through the traditional fundingmechanisms. For example, none of the recentlydeveloped second-generation road fundsappears to pay any attention to NMT or to haveNMT interests represented in the roads author-ity boards managing the funds. Given that thecommercial rationale of user involvement in fund

management is that those who benefit shouldpay, and vice versa, this is perhaps not surpris-ing. Nevertheless, it highlights a weakness of thedevice in addressing the interests of noncom-mercial stakeholders.

While the operating costs of cycling are very lowand the benefits very great where no other alter-native is affordable, the cost of bicycles is a seri-ous impediment to cycling in many countries(especially where the probability of loss by theftis high). Previous efforts to overcome this imped-iment have had mixed success but offer someuseful lessons for design of future schemes.10 Forexample:

• A World Bank–funded credit program directedto the poorer sections of the urban popula-tion of Lima met its credit targets, but washampered by the mutual distrust between staffof the Caja Municipal and the poor in theadministration of the scheme.

• An African pilot scheme giving incentives foremployers to establish credit and savingsschemes failed in situations where the cred-its were to be prefinanced by employers.

• Bicycle sale-on-credit programs have hadlimited success in pilot projects in Africa, due

BOX 9.4 STAKEHOLDER PARTICIPATION IN NONMOTORIZED TRANSPORT

DEVELOPMENT IN TANZANIA

In experiments undertaken in East Africa as part of the SSATP, different forms and roles of user par-ticipation were employed at different stages of the projects. For example, in Dar es Salaam 64user groups of about 10 people each discussed general problems with mobility and NMT andprovided information on problems, road-use behavior, potential solutions, and priorities to the plan-ners. Subsequently a general user platform of about 20 nonpolitical, nongovernmental citizens wasasked to review, articulate, and prioritize the problems mentioned by user groups; this user plat-form acted as an intermediary between the authorities and the community. Local user platformswere used to create and review plans, and propose changes; to provide control during con-struction; to mobilize resources; to organize repairs; and so on. Finally, formal user associationswith legal status and able to mobilize finances through contributions of members or users may alsohave a part to play. In the case of Dar es Salaam, a garden park was planned that is to be handedover to a user association for maintenance and protection.

Source: De Langen and Tembele 2000.

134 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

to the perceived high risk of lending to poorclients.

• In Morogoro, Tanzania, entrepreneurs whorent or lease bicycles welcomed the conceptof bicycle lease contracts, though they wereunable to prefinance enough bicycles.

• In contrast, a scheme for promotional bicy-cle sales to women and children enteringsecondary school in Morogoro, showed ahigh price elasticity of demand for bicycles,and suggested possibilities of extendingbicycle use by identifying promising targetgroups.

CONCLUSIONS: A STRATEGY FORNONMOTORIZED TRANSPORT

The major elements of a strategy for NMT shouldinclude the following:

• Clear provision for the rights, as well asresponsibilities, of pedestrians and bicyclistsin traffic law

• Formulation of a national strategy for NMTas a facilitating framework for local plans

• Explicit formulation of local plans for NMTas part of the planning procedures of munic-ipal authorities

• Provision of separate infrastructure whereappropriate (for safe movement and forsecure parking of vehicles)

• Incorporation of standards of provision forbicyclists and pedestrians in new road infra-structure design

• Focusing traffic management on improvingthe movement of people rather than ofmotorized vehicles

• Training of police to enforce the rights ofNMT in traffic priorities, as well as in acci-dent recording and prevention

• Incorporation of responsibilities for provisionfor NMT in road fund statutes and procedures

• Development of small-scale credit mecha-nisms for finance of bicycles in poor countries.

NOTES

1. Sachdeva 1998.2. De Langen and Tembele 2000.3. Sachdeva 1998.4. Men’s share in New Delhi, Accra, Lima, and

León is 100 percent, 99 percent, 84.6, and 90 per-cent, respectively.

5. Interface for Cycling Expertise 1997.6. Interface for Cycling Expertise 2000.7. Interface for Cycling Expertise 2000.8. At one of the very few places in Lima where

you can park your bike—the Catholic University—students have to show their identification beforeentering the university compound. So far, not asingle bike has been reported stolen.

9. De Langen and Tembele 2000.10. Interface for Cycling Expertise 1997.

THE ROLE OF PRICES IN URBANTRANSPORT

In most markets, prices have two main functions:to ration and allocate the use of resourcesbetween the production of different products,and to finance production. In transport in gen-eral, but in urban transport in particular, the per-formance of these functions is subject to threemajor complications:

a. The separation of responsibility for infra-structure from that of service provision

b. The pursuit of multiple objectives, particu-larly in public transport policy

c. The separation of infrastructure financingfrom charging.

In this chapter we address each of these com-plications and suggest ways in which municipalgovernments may improve their effectiveness inthe provision of transport services by develop-ing better pricing and funding arrangements.

CHARGING FOR THE USE OFINFRASTRUCTURE

Roads are “congested” when traffic volumereaches a level at which the flow of traffic is sig-nificantly impaired. In such circumstances theextra cost to society of the marginal trip is notonly the cost that the extra vehicle itself incursbut also the sum of the marginal delays that itimposes on all existing road users. Unless thecost charged to, and perceived by, each userreflects that extra cost imposed on society, therewill be an incentive to make road trips for whichthe benefit to the trip maker is lower than theextra total cost to society.1

In practice, in most countries—industrialized anddeveloping—urban roads are provided to theirusers without any direct charge. The only pay-ments from the private user to the public sup-plier that vary with the amount of road use comeindirectly in the form of taxes (primarily on fuel).Underpricing is endemic. Even in developingcountries with relatively higher taxes on fuel, taxesdo not reflect the costs of urban congestion.

URBAN TRANSPORT PRICING AND FINANCE 135

URBAN TRANSPORT PRICING AND FINANCE10Prices allocate resources and raise revenue. Urban transport pricing is

complicated by the multiplicity of objectives pursued and by the institutional

separation of road infrastructure from operations, of infrastructure pricing from

charging, and of roads from other modes of transport. In the interests of both

urban transport integration and sustainability, developing countries should move

toward prices reflecting full social costs for all modes; to a targeted approach to

subsidization reflecting strategic objectives; and to an integration of urban

transport funding.

136 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

The fact that these taxes do not cover costs hasseveral adverse effects. First, it distorts the choiceof mode in favor of road transport, particularlyprivate cars. Second, it encourages excessive useof the infrastructure (which may cause “excess”congestion). Third, because there is no direct rev-enue, it is not logically possible to use conven-tional commercial investment criteria in decidinghow much capacity should be provided. Fourth,because the revenues do not accrue to theresponsible local authority, there may be inade-quate money for proper maintenance of the exist-ing infrastructure. For all these reasons, it isdesirable to ensure that the price charged tousers at the margin covers the full social cost oftheir trips.

CONGESTION PRICINGThe concept of congestion pricing is that roadusers should pay a price that reflects the short-run marginal social cost of road use, and thathence varies according to the prevailing level ofcongestion.2 Congestion pricing has long beenadvocated by economists on the grounds that,unlike the various administrative controls whichare often used to manage traffic, it gives correctincentives over the whole range of dimensionsinvolved in travel decisions, including choice ofdestination, time of travel, mode of transport,route, and so on. Moreover, if congestion pric-ing were applied in the context of a flexible landand property market, the city would evolvetoward a more compact form, with more mixedland use, fewer resources devoted to the spreadof the road network into surrounding areas, andmore funds available for upgrading infrastruc-ture in the already urbanized area.3 Above all,unlike administrative restraint alternatives, con-gestion pricing yields revenue rather than beinga cost to the public purse (box 10.1).

For these reasons the World Bank strongly advo-cated congestion pricing in the 1975 urban trans-port policy paper (World Bank 1975), and hassubsequently examined its applicability in indus-trialized countries.4 Nevertheless, attempts tointroduce it in Kuala Lumpur and Bangkok in asso-

ciation with Bank projects foundered, as did theinitiative in Hong Kong, China, in the early 1980s.To date, the most sophisticated applicationremains that in Singapore, where an area licens-ing scheme (ALS) first introduced in 1973 has nowbeen developed into a much more sophisticatedelectronic pricing system. Instead, various surro-gate measures, such as the use of parkingrestraints and pricing policies, have been adoptedin many countries. Even in industrialized coun-tries, there is growing concern that these surro-gates have not worked well. Direct charges forurban roads have been introduced to generaterevenue in some Norwegian cities, and congestionpricing is now under serious consideration in theNetherlands and the United Kingdom.

Many of the objections raised against congestionpricing have already been overcome. Initial con-cerns about the cost and reliability of the tech-nology have been superseded by developmentsin electronics. Fears about intrusion on privacy—which contributed to undermining the Hong Kongexperiment—can be overcome through the choiceof a technology not relying on any centralizedrecord of vehicle movements. Fears about theeffects of congestion charging on lower incomegroups, which have been a significant obstacle toits political acceptability in highly motorized soci-eties,5 are much less justified in developing-coun-try cities, where most people remain dependent onpublic transport or nonmotorized modes.6 Evensome of the rapidly motorizing major cities ofEastern Europe still offer much more extensiveand frequent public transport services as an alter-native to the private car than are available in manyof the Western countries. Specific inequities thatdo arise in the shift to a more direct and efficientsystem of charging for road use can also be com-pensated by provision of case-specific subsidies,such as free or reduced-price smart cards or stick-ers during a transition period. Congestion pricingis increasingly being viewed as a concept whosetime has come.

Congestion prices can be charged with differentdegrees of precision through a variety of tech-

URBAN TRANSPORT PRICING AND FINANCE 137

niques of different degrees of technologicalsophistication and cost. While, theoretically, dif-ferent prices can be set for each link in the net-work for each time period, in practice a ratherrougher approximation may be used both forreasons of practicability of application and forpredictability of response from drivers. Threeprincipal forms of congestion pricing have beendeveloped to date:

Cordon pricing, or area licenses, can be imple-mented with simple technology, to charge for theright to access or circulate within limited geo-graphical areas, with some degree of time dif-ferentiation. The principal application has beenthe area l icensing scheme for downtownSingapore in operation between 1975 and 1998.The scheme being prepared for central Londonis of this type.

Time-dependent tolling of individual roads orroad lanes can charge for congestion on majorhighways and improve traffic flows on the affectedfacility (though not necessarily beyond it). Indeveloping countries it has been used on anumber of tunnels in Hong Kong, the Namsantunnels in Seoul, as well as on the expresswaysin Singapore.

Electronic road pricing (ERP) enables more pre-cise differentiation of charges by road, time ofuse, and type of vehicle for whatever area is cov-ered. Recent developments in intelligent trans-port system technologies make this much moreattractive, and several large test applications havebeen undertaken. In 1998 Singapore replaced itsALS by an ERP system that applies to the citycenter and some major access roads, and is grad-ually being extended as needed (box 10.2).7

BOX 10.1 THE NAMSAN TUNNELS IN SEOUL: SIMPLE ROAD PRICING REDUCES

CONGESTION AND FINANCES TRAFFIC MANAGEMENT

Traffic congestion in Seoul increased dramatically during the 1980s and early 1990s despite exten-sive construction of new urban freeway and subway lines. In 1996 the Seoul metropolitan gov-ernment commenced charging 2,000 won (US$2.20) for the Namsan #1 and #3 tunnels, two corridorswith high private vehicle use linking downtown Seoul to the southern part of the city. Chargeswere set for one- and two-occupant private vehicles (including driver) and collected in both direc-tions per entry or exit from 7:00 a.m. to 9:00 p.m. during weekdays and from 7:00 a.m. to 3:00p.m. on Saturdays. Private cars with three or more passengers, taxis, and all kinds of buses, vans,and trucks were exempted from charges, as was all traffic on Sundays and national holidays.

In the two years following commencement of the congestion pricing scheme for the Namsan #1and #3 tunnels, there was a 34 percent reduction in peak-period passenger vehicle volumes, theaverage travel speed increased by 50 percent, from 20 to 30 km/h, and the number of toll-freevehicles increased substantially in both corridors. On the alternative routes, traffic volumes increasedby up to 15 percent, but average speeds also increased as a result of improved flows at signalizedintersections linked to the Namsan corridors and increased enforcement of illegal on-street park-ing on the alternative routes.

The whole of the annual revenue from the two tunnels (equivalent to about US$15 million) goesinto a special account used exclusively for transport projects, including transport systems man-agement and transport demand management measures throughout the city.

Source: Hwang, Son, and Eom 1999.

138 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

In principle, once roads are charged at a priceequal to the marginal social cost, the decision bythe road authority as to whether and when to investin expansion of the road should depend on therelationship between congestion-pricing revenuesand the costs of the expansion.8 In practice, wherethe pricing system is only roughly applied, it maystill be necessary to undertake a cost-benefit analy-sis of proposed investments. Nevertheless, theapplication of the pricing scheme will create theproper base on which to undertake the appraisaland will have generated a revenue pool that canbe used to finance extra capacity.9

The fact that efficient congestion prices, or park-ing charges used as a proxy for them (see dis-cussion below), yield revenues in excess of thetotal direct costs of congestion is sometimes used

as an argument against them. This should not beseen as a problem but as an opportunity. Highuser prices are quite apposite to the shortage ofstreet space relative to built area in many devel-oping-country cities. The distributional implica-tions are very progressive because those who dohave cars are predominantly high-income earn-ers likely to attach the highest value to reducedtravel time and journey predictability. They mayalso be democratically viable, since only a minor-ity of the wealthier citizens would be called uponto pay the increased charges due to low car own-ership. For all of these reasons, congestion pric-ing, accompanied by use of the revenues toimprove public transport and other city ameni-ties to the benefit of the poor, can be recom-mended for the typical cash-strapped municipalityin a developing or transition economy.

BOX 10.2 ELECTRONICS IMPROVES ROAD-PRICING EFFICIENCY IN SINGAPORE

In September 1998 the government of Singapore replaced the manually enforced ALS by ERP. Itgenerally covers the same area as that for the ALS, but has been extended to work on principlessimilar to ALS, but with extension to approach and bypass roads. All vehicles are required to havean electronic in-vehicle unit (IU) that accepts credit in the form of a smartcard. Tolls are automat-ically paid when the vehicle passes under a gantry and a liquid crystal display indicates the cur-rent credit balance. Tolls do not fluctuate in relation to actual traffic volumes but are adjustedquarterly to ensure optimum traffic speeds. The system cost $200 million in Singapore dollars(US$125 million) to implement, one-half of which was for the free fitting of IUs.

The ERP system was not implemented to increase government revenue. ERP charges are gener-ally less than the corresponding ALS fees, although the ERP system’s per-pass charging principlemeans that those motorists who use the priced roads must now pay more. Overall, revenue isabout 40 percent less than that previously collected from the ALS, but electronic charging givesgreater flexibility to set charges that are just sufficient to keep the roads free of congestion.

Once new traffic patterns stabilized, weekday traffic volume entering the restricted zone droppedby 20 to 24 percent, from 271,000 vehicles per day to between 206,000 and 216,000. With thoselower traffic volumes, average traffic speeds in the zone increased from 30–35 km/h to 40–45km/h. Improvements are less clear on the three expressways in the ERP scheme, and the LandTransport Authority is reviewing ERP charges to further optimize traffic flow.

Source: PADECO 2000.

URBAN TRANSPORT PRICING AND FINANCE 139

FUEL TAX AS A SURROGATE USERPRICE For the moment, however, the tax element of thepump price of fuel is the main charge directlyassociated with road use. Fuel price is thus veryimportant in urban transport. For the privatemotorist, it is usually the only incurred cost thatis perceived in making a marginal trip. It may thusdirectly affect the amount of travel undertaken,choice of transport mode, and choice of vehicletechnology, and may less directly affect the trade-off between location and transport expenditure.

In an ideal world, the price of fuel should cover itsresource cost (the border price); maintenance ofroads and congestion costs should be chargedfor directly through highly differentiated tolls;environmental costs should be charged forthrough emissions charges; and any redistribu-tion objectives should be pursued throughnondistorting lump sum taxes.

In reality, that prescription is not achieved cur-rently in either industrialized or developing coun-tries. Most roads are open access (particularly inurban areas), and cannot be subject to highly dif-ferentiated direct charges. The technology forlocation- and pollutant-specific emissions charg-ing does not exist. Lump sum transfers or naïvedirect distributional taxes are not possible. In theabsence of direct charging mechanisms, fuel tax-ation is often looked to as a proxy for a numberof other, theoretically preferable, taxes or chargesrelating to road maintenance, congestion, andenvironmental and distributional objectives. It isthus important to examine the potential of fueltaxation to perform these various functions.

As a charge for road maintenance, fuel tax doesnot differentiate well by vehicle type, and needsto be supplemented by axle-weight charging.Similarly, as a congestion charge, fuel tax doesnot reflect the structure of congestion wellbecause fuel consumption is not sufficiently sen-sitive to vehicle speed variation. As an environ-mental charge, it is likely to be a very good proxyfor carbon emissions taxation, but many other

emissions are not proportional to fuel consump-tion and vary by fuel type, vehicle type, and extentof emissions technology employed. As a distri-butional instrument, it has the right combinationof low price elasticity and high income elasticity,and is therefore a very promising redistributor indeveloping countries.

Fuel tax does have one great advantage com-pared with taxation on vehicles (vehicle exciseduties, registration taxes, insurance charges, andso on), which does not vary with vehicle use. Thatconcerns the relationship with public transportfares, which vary directly with use (period passesare relatively unpopular in industrialized countriesbecause of the element of advance paymentinvolved). If taxation on private cars is not relatedto use, the consequence is that the cost of car usefor the marginal trip may appear very inexpensivein comparison with that of public transport. One ofthe great advantages of increasingly including pay-ment of what are presently fixed charges withinfuel taxation, or other distance-based charges,10

is that it puts private and public transport pricingstructures on a more-comparable, and hence less-distorting, basis.

Reliance on fuel taxation as a surrogate for directcharging faces some severe administrative diffi-culties. There are multiple transport fuels, used inmultiple sectors, to achieve multiple objectives.Diesel, which may merit heavy taxation for localair pollution reasons, is also the primary fuel forroad freight transport, for which reason, as an inter-mediate input rather than a final consumptiongood, it would justify a relatively low tax burden.Kerosene typically carries low taxation, or is evensubsidized, because of its use for domestic heat-ing in many countries. Unfortunately, if these fuelscarry low tax, there is a high likelihood of substi-tution of diesel for gasoline vehicles, or the adul-teration of gasoline with kerosene or diesel, eitherof which could be environmentally very damag-ing. Particularly where there is spatial differentia-tion of fuel taxation for congestion chargingreasons, fuel carrying could also be a problem ineverything except the very largest cities.

140 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

Above all, however, fuel taxation is usually anational prerogative, with revenues accruing tothe national treasury rather than to the munici-pality. For this reason it is likely to be difficult tofit it into, or coordinate it with, a city-level strat-egy, except on the basis of a countrywide agree-ment on allocation of all transport-related taxrevenues and expenditure responsibilities.

OTHER LESS-DIRECT TAXES ANDCHARGES FOR ROAD USE Other vehicle taxes may also perform a functionof allocating road costs equitably and efficientlyamong vehicle categories. In particular, annualcharges based on standard axles are widely usedto correct for the fact that road wear is much morea function of axle weights than of fuel consump-tion. Even though these taxes may be fixed cen-trally, the revenues yielded sometimes accruedirectly to a local jurisdiction.

Some taxes have a very clear allocation function.In Denmark and Hong Kong, acquisition taxesthat approximately triple the cost of cars havebeen major factors l imiting motorization(presently some 330 and 60 cars per 1,000 inhab-itants, respectively).11 The most extreme form ofthis is in Singapore, which in 1990 introduced asystem in which a rationed number of entitle-ments to own a motor vehicle (the so-calledCertificates of Entitlement) are auctioned eachmonth. In combination with preexisting taxes,this has resulted in new cars retailing at somefour or five times their world market price. Otherrelevant experiences are those of Tokyo, whereprospective car purchasers are required to pro-vide proof of having an off-street parking space;and of various Chinese cities, where entitlementsto register a motorcycle have been sold and localcirculation of motorcycles registered elsewherebanned.

Parking charges are often used in place of directroad-use charges. While they cannot deterpeople from driving through an area or reflectthe different distances or routes drivers take, theycan be varied by time and place in such a way as

to capture a significant part of the congestionexternality and encourage better spatial or tem-poral distribution of demand for movement onthe roads.12 They are already manipulated in quitesophisticated ways in many European cities, andboth the United Kingdom and France are debat-ing taxation of downtown parking spaces pro-vided by employers for their employees.13

Transferable permits for a given number of avail-able parking spaces would be a way of auction-ing the right for commuters to drive to work.

In summary, an ideal charging regime for the useof infrastructure should allocate all of the asso-ciated costs (congestion, environmental, wearand tear on the infrastructure, and so on) to theusers incurring them, as far as possible directlyin proportion to the costs imposed. That is notcompletely achievable even in principle becausethere are some costs (such as lighting, streetcleaning, and so on) that have the characteristicof public goods and hence must be allocatedmore arbitrarily. Because the imposition of thedifferent types of cost varies between vehicletypes according to a range of characteristics ofvehicles and their use, only the simultaneousapplication of a number of different chargingdevices could meet this ideal.

The design of a charging system thus needs tobe constructed from a carefully designed com-bination of different charging components. Forsimplicity’s sake, it may be desirable to limit thenumber of these components, but a number ofgeneral elements can be recommended.

• Fuel should never cost less than border pricein any circumstances.

• Where direct congestion charges are possi-ble, these should be varied by time of dayand structured according to the congestion-creating equivalents of different vehicletypes.

• Wear-and-tear costs should preferably berecouped on a basis variable both by wear-ing effect and distance traveled (standard axlekilometers). Where this is not possible, the

URBAN TRANSPORT PRICING AND FINANCE 141

wear-and-tear cost may be incorporated partlyin a fuel tax surcharge and partly in vehiclecategory–related charges (preferably alsobased on actual annual use of each vehicle).

• Marginal environmental impact costs andmarginal public costs of accidents might berecouped through local fuel tax and insur-ance surcharges.

• Any sumptuary, or “luxury” taxation, leviedon road users for general taxation (non-transport) reasons should be levied on pas-senger transport rather than freight transportin order to minimize economic distortion.

• Where diesel fuel is taxed at a low rate, somecompensating tax should be imposed onthe procurement and use of diesel-poweredlight passenger vehicles to minimize adverseenvironmental consequences.

PUBLIC TRANSPORT PRICING AND FINANCE

Setting public transport prices and raising thenecessary finance, raises problems because ofthe multiple objectives faced by decisionmakers.The primary objective of public transport pricingis to generate revenue that can ensure an effi-cient and adequate supply of public transportservice. Public transport pricing may also beexpected to contribute to the reduction of con-gestion and environmental impact of road traf-fic, efficient coordination between publictransport modes, and the reduction of poverty.It is commonly argued that if urban public trans-port is to satisfy these latter objectives, it cannotbe expected to cover its full costs. Urban publictransport is consequently subsidized in manymajor cities in industrialized countries.

Similar policies have traditionally been appliedin the transition economies and some postcolo-nial developing countries, but many of thesecountries are no longer in a position to fund suchpolicies, and their public transport sectors arefacing decline because of cash starvation. Thissection considers how to establish a basis for set-

ting policies on pricing and cost recovery in urbanpublic transport that best reconciles the multi-ple objectives.

EFFICIENT SUPPLY There are two important aspects of supply effi-ciency. First, it is necessary to provide the mostbeneficial range of services with the resourcesavailable—“doing the right thing.” Second, it isnecessary to supply the required services at theleast-possible cost—“doing the thing right.”Neither is simple and neither is well achieved inmost developing countries.

The difficulty in doing the right thing is that theremay be legitimate reasons, both on equity andefficiency grounds, why a pure market outcomewould not be optimal. For these reasons, munic-ipal governments may intervene to define boththe services to be provided and the fares to becharged. In the absence of the capability to makeany more sophisticated calculation, this typicallytakes the form of setting flat fares for a definednetwork of basic services. Public sector monop-olies are often protected because of the difficultyof enforcing a flat-fare policy with multiple oper-ators.

Such monolithic simplicity is rarely optimal. Transportusers have different preferences (partly as a result ofdifferences in income). It may therefore be possi-ble to increase consumer welfare while increasingprofitability by supplying different products to dif-ferent segments of the market. In urban bus trans-port, this means the provision of “premium”services—such as express services or air-conditionedservices—at premium prices. Price discrimination,in this sense, may benefit all users. Operators shouldtherefore be encouraged to examine such revenue-enhancing strategies as an alternative to relianceon subsidy. This strategy has already been widelyadopted as a means of maintaining urban bus serv-ice in Seoul, Buenos Aires, Bangkok, and many otherdeveloping-country cities.

Doing the thing right is what commercial com-petition usually ensures, since the threat of bank-

142 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

ruptcy is a powerful stimulant to internal effi-ciency. The possibility of subsidy weakens thatincentive. Particularly in parastatals, subsidiesstrengthen the likelihood that organized laborwill appropriate part of it to support better wagesand working conditions than a competitivemarket would be able to support. In the UnitedKingdom it was estimated that over one-half ofsubsidies aimed at lowering fares or improvingservice quality actually “leaked” into benefits formanagement or workers or were lost throughreduced efficiency of operation.14 Exploringmeasures to reduce subsidy requirementsthrough improved efficiency and reduced costsis thus the first step to take in formulating apublic transport subsidy strategy.

Much of the argument for transport deregulationand privatization (see chapter 7) derives from aneed to reduce subsidies. Attempts to increaseefficiency of operation through the introductionof competitive pressures within the sector mayenable lower prices to be charged withoutrecourse to subsidy. The implication is that, forany level of cost recovery lower than 100 percent,subsidies should be specifically targeted at theobjectives sought and should be embodied incompetitively tendered service contracts.

The macroeconomic effects of urban transportsubsidies also need to be considered. In the shortrun, they may reduce the general price index iftransport has a disproportionately heavy weightin the index. Particularly where urban transportis a significant item in the consumption pattern ofstrongly unionized labor groups, there may besome associated relief from wage pressure. Inthe longer run, however, it is inevitable that theinflationary pressure will spill over to other itemsthat will have to be taxed to finance the subsidy,o r th rough inc reased d i rec t i ncometaxation–inducing compensatory wage claims.Deficit budgeting has similar inflationary conse-quences. The general World Bank position is thatsubsidy is the wrong tool to deal with inflation;there is nothing special about the transport sectorto vitiate this view.

THE IMPACTS ON ROAD CONGESTIONIn industrialized countries it is commonly arguedthat public transport should be subsidized inorder to entice trip makers out of private cars andhence reduce road congestion. Where road con-gestion is concentrated in particular locations orat particular times of the day, this implies a highlydifferentiated structure of subsidy. Where roadcongestion is systemwide and pervasive through-out the day, general public transport subsidy maybe called for on this argument.

Despite the superficial attractiveness of this lineof reasoning, there are several considerationsthat counsel caution in the use of public trans-port subsidy as a countercongestion instrument.First, there is the supply efficiency impact dis-cussed earlier. Only if the benefits resulting froma more efficient modal split were greater thanthe disbenefit of any reduction of supply effi-ciency associated with subsidy would there be anet benefit overall.15

Second, there is a problem of targeting. Wherecongestion is limited in time or spatial extent, itbecomes increasingly difficult to target publictransport subsidies as a response to the prob-lem. Highly peaked road congestion calls forpeak-hour public transport subsidies, which mayhave the perverse side effect of shifting publictransport demand from off-peak to peak peri-ods. For spatially variable congestion, differen-tial subsidy policy by route may be extremelydifficult to structure even if competitively ten-dered service franchising allows each route tobe treated differently.

Third, there is a problem of fiscal cost. Where thebasic purpose of the subsidy is to divert trafficfrom the private auto to public transport, the idealsituation would require very high cross-elasticityof demand for auto use with respect to publictransport pricing. Empirical evidence (albeitmostly from industrialized countries) suggeststhat in fact this cross-elasticity is very low (per-haps 0.1) in the short run. While it may be some-what higher in the long run due to effects on car

URBAN TRANSPORT PRICING AND FINANCE 143

ownership, it would still appear to be an instru-ment of relatively weak leverage.

Finally, there is a problem of perverse land-useeffects. Public transport subsidy to countervailsystematic undercharging for private transportwill mean that all transport is subsidized. This willtend to generate excessive travel and inappro-priately sprawling land use. Measures to redressinadequacies in the system of charges for privatetransport should always be a condition attachedto the financing of “second-best” subsidies.

PUBLIC TRANSPORT COORDINATIONAllocation of traffic between competing modeswithin an urban transport system would be effi-cient if prices were everywhere equal to short-run marginal social costs. The general prescriptionfor covering fixed costs where these will not berecouped by short-run marginal cost pricing is tominimize distortion by loading the fixed costs

most heavily on those parts of the market forwhich the demand is least likely to be affectedby a price increase. This is known as the “Ramseypricing rule” (box 10.3). In multimodal urban trans-port systems, which are, as a whole, in financialequilibrium, this may imply cross-subsidy betweenmodes, particularly when the modes have dif-ferent cost structures. The application of thisapproach to urban public transport pricingrequires consideration, not only of the implica-tions of the relative congestion externalities ofalternative modes but also of their different coststructures.

Where two modes are operated within a singleorganization, and entry is regulated, it is possibleto reconcile different levels of cost recovery forthe two with any overall cost-recovery requirement(including breakeven overall). It is quite commonfor this to occur in metropolitan public transportundertakings (though it should be remarked, in

BOX 10.3 RECOVERING FIXED COSTS: THE RAMSEY PRICING RULES

The Ramsey inverse elasticity mark-up rule concerns the allocation of fixed costs between alter-native products within a production agency. If a multimodal transport system is viewed as a singlesupply agency, the Ramsey pricing rule can be interpreted as a rule about the allocation of system-fixed costs between modes. For example, in typical bus operations, over 90 percent of costs varywith respect to either the number of vehicles employed or the number of bus kilometers run.Short-run marginal cost pricing would nearly cover full cost. The same is not the case for rail sys-tems, where typically only 50 or 60 percent of costs are directly related to the service provided. Insuch conditions, the most efficient outcome may involve different levels of total cost coverageby the modes and transfers between them.

If cross-elasticities between alternative transport forms are zero, the objective of efficiently cov-ering total costs will be achieved by adding mark-ups to the marginal social cost in proportionto the reciprocal of the price elasticity of demand. The rule then becomes rather more complex.Consider, for example, the ratio of peak and off-peak prices. If the peak demand were completelyinelastic and the off-peak demand infinitely elastic, the whole of the burden would fall on thepeak price. More realistically, some users may shift from peak to off-peak as relative prices change.With nonzero cross-elasticities, the ratio of the mark-ups of different service should take intoaccount the cross-elasticity as well as the own-price elasticities.

Source: Authors.

144 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

passing, that where it does occur, it is oftenregarded as a form of illegitimate cross-subsidy).

However, when the modes are operating in com-petition with each other, attempting to apply theinverse elasticity rule creates a conflict betweenthe interests of pricing for allocative efficiencyand those of pricing for commercial viability.Competitive operation of such mixed-mode sys-tems on a purely commercial basis will result ina suboptimal level of patronage for those modeswith the greatest proportion of fixed cost (emptytrains and overfull buses). It will also affect thestructure of services provided. For example, com-mercial operation of buses and metros within anurban area may lead to a smaller amount ofrestructuring of bus services to take advantageof the complementarity of the modes and a largeramount of trunk service by buses than would beoptimal. Recent experience in Rio de Janeiro sug-gests that concessions of urban rail operationsto the private sector may make the achievementof fare integration more difficult.

Not all modal coordination considerations militatein favor of public transport subsidy. Because con-gestion must, by definition, be greater at peakthan at off-peak hours, the marginal social costof the peak period private car user must alwaysexceed the average cost. The marginal social costof peak-hour public transport may also exceedits average cost, because peak service requiresextra vehicle capacity costs and may impose highmarginal labor costs (resulting from labor agree-ments requiring payment for a minimum numberof hours per day exceeding the number of peakhours, and premiums for any splitting of workingshifts). The marginal social cost of peak publictransport provision thus depends not only on thelevel of congestion but also on the disparitybetween peak and off-peak demand levels andthe industrial relations context in which the serv-ice is being supplied. For cities with relatively lowlevels of road congestion but high public trans-port peak-to-off-peak demand ratios, optimalpeak public transport subsidies may be small (oreven, conceivably, negative).

EQUITYIn many cities public transport fares are controlledin an attempt to maintain an affordable serviceto the poor who have no alternative form of travel.While in principle lump sum transfers would be apreferable way to redistribute income, in prac-tice they are rarely feasible politically. Increases inpublic transport fares are therefore politically sen-sitive. (Five people died in riots following a fareincrease in Guatemala in April 2000.) The ques-tion is thus not whether the objective of main-taining affordable public transport service isimportant and desirable, but whether it can bepractically and cost-effectively achieved with thepolicies typically adopted for the purpose. Themain characteristics to review in this context arethe general fare level, fare structures, fare con-cessions, and fare discrimination.

Fare levels. In most low- and middle-incomedeveloping countries, the average income ofpublic transport users is below that of motorizedprivate transport users, and below the overallaverage income. If fares are set below cost insuch circumstances and operating deficits cov-ered either from a progressive income tax or atax on private motoring, the distributional(poverty alleviation) impact will be positive. If,however, the control on fares is not directlyfunded, the long-term effect will be to reducefirst the quality and eventually the quantity ofpublic transport available. Whether the pooractually benefit from price controls then dependson the tradeoff between less expensive fares andpoorer service. There is considerable evidence,even in relatively poor countries, such as theKyrgyz Republic, that the poor are willing to paymore for a better service than that provided atexisting controlled fares. In the extreme, the poorget no benefit at all from setting very low faresif that causes supply to disappear altogether.Decisions on the control of fares should thus betaken in the context of an assessment of theeffects of the control on the quality of serviceand in the context that the continuation of anappropriate level of service can continue to befinanced.

URBAN TRANSPORT PRICING AND FINANCE 145

Fare structure. Flat fares are frequently adoptedthroughout a municipal or conurbation area inthe belief that this is equitable. Where longerwork journeys have been forced on the poor,either through racial discrimination in housing(as in South Africa) or through market-unawarehousing planning (as in many former commandeconomies), flat fares may indeed compensatefor other forms of discrimination against the poor.The danger, however, is that in larger cities, a flatfare may need to be set at such a high level thatit encourages shorter-distance travelers to seekalternative modes, and hence to undermine any“within-mode” cross-subsidy. It is therefore rec-ommended that such spatial distribution issuesbe handled through specifically designed serv-ices and fares, rather than through a general flat-fare system.

Designing a service and fare structure that is bothequitable and efficient is particularly difficult inmultimodal systems. There may be an efficiencycase for structuring bus services to feed high-capacity trunk rail systems and for allowing thebuses to cross-subsidize the rail system. In manycases, however, the average bus user is poorerthan the average rail user, so what may look effi-cient from a coordination viewpoint involves thepoor subsidizing the rich. The lack of multimodalticketing accentuates this problem The moral isthat wherever explicit attempts are made tosecure modal coordination within urban publictransport systems, it should be in the context ofan integrated fares and charges strategy takingparticular cognizance of the effects on the poor.

Fare reductions or exemptions. The distributionof poverty-oriented subsidies should ideally beexplicitly related to the income levels of serviceusers. There may be some vulnerable categoriesof passenger (schoolchildren and senior citizensnot in receipt of a full income) that can be easilydistinguished and for which charging a lower priceis possible. The advantage of discrimination onthis basis is that it can be applied to all servicesand can even be nuanced (for example, to excluderetired people with incomes above some mini-

mum level). In many countries, and particularly inthe transitional economies, there is a pervasivetradition of providing free or reduced-fare trans-port for a wide range of public servants. This raisestwo major problems. First, the categories inreceipt of free fares are often not the most needy,so that the redistribution effects of supportingthe concessions by internal cross-subsidy are per-verse. Second, where there are a large propor-tion of non-fare-paying passengers, it becomesmore difficult to enforce payment by those whoare supposed to pay. Explicit consideration of theextent and justification of fare exemptions andthe extent and means of combating fare evasionshould therefore be a sine qua non of any fare-determination process. Moreover, where reducedfares are mandated for the support of other sectorpolicies (health, social security, policing, and soon), the costs of the subsidies should be chargeddirectly to the other sector budgets.

Fare discrimination. Where residential locationsare highly segregated by income group, specificroutes may also be identified for subsidy onincome-distribution grounds. For example, atone stage a fare differential existed between Lines1 and 2 of the metro in Santiago, Chile, with theline serving the lower-income population charg-ing lower fares per kilometer. Less well targetedapproaches may have perverse effects. For exam-ple, the provision of minimum levels of networkdensity and frequency throughout an urban areaat the standard fare level will tend to providehigher levels of service or lower price to areaswith low residential density than would be com-mercially sustainable. In middle-income coun-tries, these areas are areas of high incomes andhigh car ownership. If such a policy were pursuedwithin a system covering its costs in total, theeffect would be a perverse cross-subsidy fromthe poor to the rich. Maintaining specified net-work density or frequency by external subsidyhas other disadvantages. In particular, there willbe a systematic tendency to overprovision,because both those who receive the subsidizedservice and those who supply it develop a vestedinterest in high levels of provision.

146 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

THE DYNAMICS OF POLICY REFORM:MAKING SUSTAINABILITY PARAMOUNTThere is a close relationship between charges forinfrastructure and charges for public transportservice. If road use were efficiently priced, therewould be no economic second-best case forpublic transport subsidy. Logically, this suggeststhe need not only for action to secure optimalpricing for both infrastructure and services butalso for careful attention to the dynamics ofchange. Eliminating public transport subsidy inthe absence of efficient infrastructure chargesmight simply further distort the choice of trans-port mode. The counsel of perfection would thusbe to link any move to improved cost recoveryfor public transport to the progress made in intro-ducing efficient road pricing.

Within the context of political reality, however,that may be a difficult prescription to observe,particularly where, because of fiscal collapse, theshort-term alternative to fare increases is not con-tinued subsidy but loss of service. Generating anadequate cash flow is the most serious problemfor many public transport operators in develop-ing countries. Attempts to control fares at uncom-mercial levels in the absence of a secure fiscalfoundation threatens sustainability of service, notonly for parastatals (as in many cities in WestAfrica) but also for franchised private operators (asin Jamaica). Without a secure financial basis, it isimpossible to achieve the other objectives dis-cussed. For that reason it is suggested that, whileincreased cost recovery through the fareboxshould not be the primary concern of public trans-port policy per se, the maintenance of a soundfinancial basis should be at the heart of the pric-ing policy process.

This is not always well understood. In transitionaleconomies, where parastatal operators are notaccustomed to commercial accounting princi-ples, it may be particularly difficult to ensure thatproper provision is made for the financing ofdepreciation. Even where some of the capitalinvestment is sunk in long-lived assets that arefixed or cannot be profitably resold or devoted to

any other use, system sustainability still requiresenough revenue to cover costs of operation,maintenance, and replacement of rolling stock.

URBAN TRANSPORT FINANCING

The financing of municipal transport is complexand difficult both because of the separation ofroad infrastructure from operations and becauseof the multiple objectives that public authoritiesare pursuing in urban transport policy. In this sec-tion we first consider the problems of financingpublic sector infrastructure, then examine thepossibilities of private sector participation in infra-structure finance, and finally consider the insti-tutional possibilities of mobilizing the pricingdevices that have been discussed above to solvethe financing problem.

PUBLIC FINANCING OF INFRASTRUCTURE The sources of funding for public sector invest-ments may include transfers from central gov-ernment, local borrowing, local taxation, andservice charging. Public expenditures on urbantransport in capital cities may be fully financed bythe central government. More generally, andalmost exclusively in noncapital cities, the mainresponsibility for finance will rest with the regionalor municipal government with some degree ofcounterpart funding from the central government.That sharing of financial responsibility can induceoverinvestment in infrastructure if it takes the formof automatic counterpart funding of locally gen-erated schemes. Projects should therefore be sub-ject to some formal economic appraisal ofinvestments as a condition for cost sharing.

Local borrowing may be secured against generalmunicipal revenues or occasionally against tollrevenue. Some municipalities are at least as cred-itworthy as their national governments and may beable to issue bonds of their own. That is not usu-ally the case for smaller cities, however, and bor-rowing may have to be by government on theirbehalf. In this context it is essential that all claims

URBAN TRANSPORT PRICING AND FINANCE 147

for capital resources be based on, and assessedaccording to, a common investment appraisal cri-terion and that the municipality should meet theservicing costs of its borrowing.

Local taxing capability is also very limited in manycountries, often being restricted to property tax-ation and various minor licensing revenues. Inmany developing countries, sales tax and incometax are central government prerogatives. Otherlocal taxes, or taxlike impositions, are sometimesused to supplement user charges for transportservices. The French “versement transport” is anearmarked payroll tax. The “vale-transporte” inBrazil is an obligation imposed on employers tofinance part of the commuting costs of theiremployees.

On the principle that “he who benefits shouldpay,” it may seem desirable to capture in tax rev-enues part of the benefit due to infrastructureinvestment accruing to local residents or busi-nesses, and to use these funds to finance invest-ments. A “betterment” tax, which appropriatespart of the gain in the value of properties result-ing from infrastructure investments, does this inprinciple but is difficult to assess and apply inpractice. It is also collected after the investmentrather than in advance. Development chargesand infrastructure contributions, when assessedsystematically, can provide for the extension ofinfrastructure as cities expand, but can rely onthe existence of very strong development con-trols, which are often lacking in developing coun-tries. The national government of Japan hasissued administrative guidance on integrated railand new town development mandating contri-butions of the land developer to the rail enter-prises with several provisions to be included intheir agreement.16

Direct charges for service from users of a partic-ular piece of locally funded infrastructure is nor-mally treated as a trading revenue that accruesautomatically to the supplying authority—butthese direct charges for service are very rare.Indirect charges, such as fuel taxation, usually

accrue to the central government. A critical ques-tion is whether congestion charges are to betreated as a charge for service or a tax. One ofthe reasons that road pricing has made so littleprogress in industrialized countries is that it isoften legally classified as a tax, accruing to thecentral treasury. The British Parliament hasrecently passed a law to allow municipal author-ities to keep road-charging revenues, and hencegives municipalities an incentive to introducecongestion charges.

MOBILIZING PRIVATE FINANCE Where public transport service is franchised tothe private sector, the financing of vehicles andtheir supporting infrastructure normally becomesthe responsibility of the franchisee. The main dif-ficulty in this context is that unless the contractsare well defined, and of reasonable duration, itmay be difficult for a private operator to financethe necessary vehicles. For example, one of themajor difficulties confronting the introduction ofcompetitive tendering of franchises in Uzbekistanis that in the absence of any history of contract-ing, neither borrowing for nor leasing of vehiclesis possible.

Recent experience in some Latin American andEast Asian countries has shown that privatefinance can also be mobilized for urban trans-port infrastructure through concession arrange-ments. Urban expressways have been totallyprivately financed in Argentina, Malaysia, andThailand, and the first urban metro in Bangkokwas privately financed. In other cases, such as theconcessioning of the existing suburban railwaysin Argentina, where prices were fairly tightly con-strained, public contribution was necessary tomake private financing of new investment viable.

The need for public sector contribution is not areason for forgoing private participation, but itdoes emphasize the need to establish both theprinciples and procedures through which publiccontributions should be appraised. Essentially, ifthe private concessionaire is able to exploit con-sumer surplus of users of the new infrastructure,

148 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

the public sector should require that the value ofexternal and nonuser benefits be sufficient to jus-tify the required contribution on its normal eval-uation conventions. The danger is that thegovernment will become committed without anyclear understanding of the costs and benefits.

One of the major impediments to private sectorinitiative in urban transport infrastructure is theproblem of acquiring many fragmented parcelsof land and assembling them into a linear right-of-way. For “greenfield” developments, trans-port infrastructure may so improve the value ofadjoining land that it should be in the interestsof landowners to facilitate infrastructure con-struction. This potential gain has been effectivelyexploited in a number of urban rail systems inJapan through a process of land assembly andreparceling (box 2.3 in chapter 2).17 The codifi-cation of a basis for such private sector initiativesis worth exploring as a means of facilitating pri-vate sector investment in large-city transport infra-structure. In areas of existing development,however, that goal is more difficult to achieve.For this reason the public sector usually uses itspowers of eminent domain to assemble and pro-vide the right-of-way, as in the cases of the pri-vate investments in new metro systems in bothBangkok and Manila.

URBAN TRANSPORT FUNDINGIt is widely agreed that urban transport planningand operation should be an integral part of urbanstrategy. It is even more obvious that the activi-ties of the different modes and functions withinthe transport sector should be well integrated.By implication that requires consistency of finan-cial arrangements within an overall strategicframework. In most countries, however, no morethan lip service is paid to that prescription. Wetherefore need to explore how financing arrange-ments might be structured to secure more effec-tive integration both within the sector andbetween sectors.

The usual institutional context for this has twocomplicating characteristics. First, responsibility

for public transport is increasingly being decen-tralized to the cities in the absence of an ade-quate local fiscal base. This means either that thesector has to be financially self-supporting or thatit must be dependent on intergovernmentaltransfers. Second, the supply of transport facili-ties and services is typically very fragmented,both between private and public sectors andfunctionally within the public sector itself. Thismeans that the provision of service usuallydepends on separate and independent financ-ing arrangements for the modes.

Intergovernmental transfersWith respect to intergovernmental transfers, amechanism is needed that channels finance in away that neither distorts the allocation ofresources between alternative modes or instru-ments nor weakens the incentives to efficientoperation of the individual modes. Unfortunately,the two most commonly used transfer mecha-nisms—counterpart funding of infrastructureinvestments and direct subsidy of bus opera-tions—fail these tests.

Counterpart funding by national governmentsof urban transport infrastructure investment, par-ticularly in capital cities, is common, while cur-rent account support to the municipalities fortransport purposes is rare. The effect of this is tobias the cities in the allocation of resources attheir disposal in whatever way attracts the mostgenerous counterpart funding. The partiality ofcities for road investment in many countries, andeven for heavy rail investments in some coun-tries, often derives from the high degree of coun-terpart funding that such investments attract. Itmay also undermine collaboration, becauseauthorities see more benefit in competing for the“cheap” counterpart funds than in collaboratingto develop effective programs.

Three devices can contribute to limit or eliminatethis rent-seeking behavior. First, allocation offunds from central to local government shouldbe based on the quality of the project. Second,all such investments in conurbations may be

URBAN TRANSPORT PRICING AND FINANCE 149

required to be in conformity with an agreed-onconurbation development plan. Third, supportcan be channeled through block grant arrange-ments that leave the cities free and responsibleto decide how central government contributionsare best used.

Subsidization of public transport operations bythe central government is typically channeledthrough deficit funding of publicly owned trans-port companies. This discourages the search bymunicipalities for more efficient supply mecha-nisms, and has ultimately failed to provide a sus-tainable basis for public transport operations inmost of the developing and transitional economieswhere it has been applied. To avoid this it is nec-essary to replace the funding of publicly ownedsupply agencies by the provision of untied fundswith which the municipalities can procure services,in whatever way they find it most effective. In somecases where municipal government has been cap-tured by the supply agencies or their employees,it may be necessary to link the provision of suchfunds with requirements for the introduction of acompetitive procurement process.

Intrasectoral coordinationThe pricing principles set out earlier clearly donot result in balanced budgets for each and everyurban transport mode. In the case of congestedurban roads, short-run marginal cost pricingwould yield a surplus over road maintenance andoperation costs. While in many sectors this wouldjustify increasing capacity so that short-run costsfell, that may not be possible in the case of urbanroads because of the high land and environ-mental costs of increasing the capacity. In con-trast, setting public transport fares below cost inorder to redistribute income, to act as a coun-tervailing distortion to the underpricing of roadspace, or to affect modal split for environmentalreasons would all leave a deficit in public trans-port financing. This might be implemented in acompetitive regime through negative price fran-chising of transport services. Even within thepublic transport sector, an optimum pricing struc-ture might require different cost-recovery rates

for road- and rail-based systems because of dif-ferences in cost structures between modes andindivisibilities in supply. In all these cases, a sys-temwide approach to urban transport financemakes more sense than a strict financial balancefor individual modes or suppliers.

Political attitudes to cross-subsidy are somewhatparadoxical. Within the bus sector, cross-subsidyof unprofitable routes or schedules by profitableones is often viewed as a good thing and the lossof that capability to be the major disadvantage ofcompetitive regimes. Yet, at the same time, thereis great resistance to the transfer of surplusesfrom one mode to another. Road pricing is fre-quently opposed because the benefits of the sur-plus are not likely to be returned to those whopay the charges. The future of congestion charg-ing will depend largely on the political accept-ability of the proposed uses for the large revenuesthat will be generated.

Public opinion on this appears to be changing inthe industrialized countries. The emerging pro-fessional consensus that “net revenues from con-gestion charging (after some reduction of otherroad-user taxes in cases where they are alreadyhigh) should be devoted largely to transportimprovements” certainly reflects the results ofpublic opinion surveys. European discussions usu-ally give particular emphasis to use of a signifi-cant portion of the revenues for support to publictransport, partly on grounds that drivers in con-gested areas benefit directly through reduced pri-vate travel time and cost from the decisions ofothers to use public rather than private transport.Various formulas have been proposed to com-bine transport service objectives with social andfiscal desiderata so that the benefits of the schemewould be distributed as fairly as possible andwould be perceived as being fair.18 The emer-gence of such a view could be very significant ininfluencing policies in developing countries.

Integrated urban transport finances These considerations suggest the need for amechanism to integrate urban transport

150 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

finances, whether secured from users, from localtaxes, or from intergovernmental transfers, andto ensure their efficient allocation among uses.For smaller, unitary cities, an existing all-pur-pose administration might be quite adequatewithout the creation of any special new institu-tion. Many Chinese cities appear to operate verysatisfactorily this way. In these circumstancesthere may already be sufficient fungibility offunds both within the transport sector andbetween sectors to serve the area well. But thelarger the city size and the more complex itsjurisdictional and functional organization, themore likely it is that a “ring fenced” fund andauthority will be beneficial.

One straightforward solution for larger conurba-tions is the creation of an urban transport fundinto which all local transport trading profits, trans-port-related intergovernmental transfers, or localtax allocations should be paid, and from whichall local public sector transport expendituresshould be financed. The creation of such a funddoes not depend on any specific tax source beingearmarked for transport, although it would be anessential part of such a scheme that local con-gestion charges be treated as municipal tradingrevenues and not as general tax revenues.19 Theessential features to be sought in an arrange-ment would thus be the pooling of all availabletransport revenues and their allocation betweenuses on the basis of the contribution made to theoverall municipal development objective.

Neither does the creation of such a fund requireany unique local political structure. While thefunds would need to be administered by a pro-fessional urban transport executive, there aremany possible structures for the strategic con-trol of the executive. Where there is a strong localdemocratic process with competent administra-tion, it might be directed by the local politicalauthority (as is being considered in Buenos Aires);in multiple-jurisdiction authorities, it might beresponsible to a joint committee of contiguousauthorities (as is being developed in many large,multijurisdictional Brazilian conurbations). It might

be administered by a “special purpose district,”as is common in the United States. The control-ling authority might even include direct user rep-resentation, as is commonly advocated fornational road funds. These alternatives are con-sidered in more detail in chapter 11.

CONCLUSIONS: A STRATEGY FORURBAN TRANSPORT PRICING ANDFINANCING

The essence of the urban transport pricing andfinancing issue is that in a system where themodes are highly interactive and policy objec-tives are complex, the separation of responsibil-ities for road infrastructure from operations, forinfrastructure provision from infrastructure charg-ing, and for roads from that of other modes cre-ates significant policy distortions. The irony is thatwith such a heavy demand for road space, andsuch palpable undercharging for its use, citiesare short of financial resources to support theinvestments and the modes of transport that canbest contribute to the relief of urban transportproblems. In the interests of urban transport inte-gration and sustainability, developing countriesshould move toward prices reflecting full socialcosts for all modes, to a targeted approach tosubsidization reflecting strategic objectives, andto an integration of urban transport funding. Atthe same time, however, it is desirable to retaininstitutions and objectives for the individualmodes or components that give high incentiveto operational efficiency and cost-effectiveness.A monolithic public sector monopoly is thus notthe solution.

The components of a strategy to achieve this log-ical integration include:

On charging for road infrastructure• Vehicular users of congested urban road

space should be charged a price at leastequal to the short-run marginal cost of use.

• Cordon pricing and tolling of specific roadsmay make an interim contribution, but the

URBAN TRANSPORT PRICING AND FINANCE 151

long-term solution must lie in a more thor-ough application of congestion charges.

• Fuel tax, although a poor surrogate for con-gestion or road maintenance impact pric-ing, should in the absence of d i rectcharging be structured along with vehiclelicense duties to give the best availableproxy.

• Taxes on different fuels should be structuredto reflect their relative contributions to urbanair pollution, again in conjunction with thestructuring of vehicle license duties.

• Parking charges, although a poor proxy forcongestion charges, should always cover thefull opportunity cost of land used for park-ing.

• Where parking policy is the only availableproxy for efficient pricing, controls need tocover all forms of parking space (includingthat provided privately by employers foremployees), and should be designed tosecure a level of restraint equivalent to thatof efficient prices.

• All road congestion charges, or fuel tax sur-charges operating as a proxy for them,should accrue to the municipal or metro-politan authority, and not to the central gov-ernment treasury.

On public transport pricing and finance• Pricing principles for public transport modes

should be determined within an integratedurban strategy and should reflect the extentto which road infrastructure is adequatelycharged.

• Given the high level of interaction betweenmodes, and the prevalent undercharging ofroad use, no absolute value should beascribed to covering all costs from fares,either for public transport as a whole or forindividual modes.

• Transfers between roads and public trans-port services, and between modes of publictransport, are potentially consistent with opti-mal pricing strategies.

• In the interests of efficient service supply,transport operators should operate com-

petitively with purely commercial objectives,with financial transfers achieved through con-tracts between municipal authorities andoperators for the supply of services.

• Any noncommercial objectives imposed onoperators should be compensated directlyand transparently, where appropriate by non-transport line agencies in whose intereststhey are imposed.

• In the absence of appropriate contractingor other support mechanisms, the sustain-ability of public transport service should beparamount and generally have precedenceover traditional price regulation arrange-ments.

On urban transport financing• Given the degree of interaction between

modes, urban transport financial resourcesshould be pooled within an urban transportfund administered by the strategic transportauthority at the municipal or metropolitanlevel.

• Intergovernmental transfers should normallybe made to the fund and should be struc-tured in such a way as to avoid distorting theefficient allocation of resources within thetransport sector at the local level.

• Private sector financing for transport infra-structure should be raised through com-petitive tendering of concessions that maybe supported by public contributions as longas these have been subject to proper cost-benefit analysis.

• When allocating funds to urban transport,the relationship between transport policyand other sector policies, in particular hous-ing, should be borne in mind.

NOTES

1. That does not mean that the optimum wouldbe a situation without any congestion at all. Inurban areas, where the cost of extra road capac-ity is usually very high, some people may be will-ing to pay a surcharge over their own direct costsfor the benefits that they receive from making a

152 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

trip, but the benefits of eliminating congestionare not sufficient to justify the high costs of theextra capacity.

2. Mohring 1999.3. Wheaton 1996.4. Hau 1992b.5. See, for example, Richardson and Bae 1998.6. Equity concerns have not been a major con-

cern in Singapore’s repeated raising of taxes andcharges on motor vehicle ownership and use,and family budget surveys make it clear that thevast majority of these taxes and charges are paidby the 30 percent of the households with thehighest incomes.

7. See Willoughby 2000b.8. Hau 1992a.9. Roth 1996.

10. For a discussion of distance-based charges,see Litman 1999.11. Other countries, too, have sometimes justi-fied new taxes on car acquisition as measures toration access to road space, but the relatively lowrates of the tax make it harder to identify specific

impact. For the case of the United Kingdom, seeNewbery 1990.12. Verhoef, Nijkamp, and Rietveld 1995.13. Conseil National des Transports 1999; andDepartment of Environment, Transport, and theRegions 1998.14. TRL 1985.15. Given widespread evidence that demandfor public transport is more elastic with respectto level of service—particularly network densityand frequency—than to price, policies aimed atsecuring high levels of service quality may wellbe more effective than those directed at prices.16. PADECO 2000.17. PADECO 2000.18. Small 1992.19. Without such a provision, municipalities areinevitably resistant to the introduction of directcharges on local road users. The recent resur-gence of interest in congestion charging in theUnited Kingdom has resulted partly from thepassing of legislation to allow local authoritiesto keep any road congestion charge revenues.

THE IMPORTANCE OFINSTITUTIONS

This review shows the heterogeneity and techni-cal complexity of urban transport issues, as wellas their political sensitivity. At the same time, itemphasizes the need for strategy integration bothwithin the urban transport sector and between thesector and the rest of the urban developmentprocess to which it contributes. This is an enor-mous institutional as well as intellectual task.

Urban transport strategy operates at three levels:

Strategy for the city is the concern of national andregional governments, which have the responsibil-ity for formulating regional development policy, forallocating intergovernmental funding transfers, andfor establishing the legal framework within whichlower-level authorities and agencies operate.

Strategy of the city is the concern of municipalauthorities, which have the responsibility fordetermining their own internal priorities, sup-plementing the resources available from localsources, and allocating the resources at their dis-posal to achieve city objectives. It is also the con-cern of citizens who may not be well heard orrepresented through the local political process.

Strategy in the city is the concern of imple-menting agencies, both private and public sector,who have the responsibility for performing thetasks, or supplying the services, attributed tothem, but who have some degree of technicalautonomy in undertaking these duties.

Table 11.1 lists the functions commonly associ-ated with each level of strategy making and thewide range of agencies that are involved. Forurban transport strategy to be effective, each ofthese three levels must have a technical compe-tence to perform their tasks, and the levels ofstrategy must be consistently aligned and effi-ciently coordinated. Many of the failures toachieve desired objectives (such as safety, envi-ronmental protection, and public transport inte-gration) or to implement policy instrumentseffectively (for example, traffic management) thathave been observed earlier can be attributed, atleast in part, to institutional weaknesses. In manycases collaboration is necessary between insti-tutions or actors operating under different incen-tives and having no inducement to act in concertwith each other. Cities that have failed to findacceptable institutional mechanisms have alsofrequently failed to address the problems ofincreasing road congestion, environmental dete-rioration, and the decline of public transport.

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 153

STRENGTHENING URBANTRANSPORT INSTITUTIONS 11Institutional weaknesses are the sources of many observed failures in urban

transport in developing countries. At the municipal level, institutional structures

for transport are weak and inadequately staffed. The need to integrate policies

both within the transport sector and between transport and other aspects of

urban development calls for the development of institutions that minimize

jurisdictional and functional impediments to policy integration and allow for

extension of the role of the private sector within an integrated strategy.

154 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

MAJOR INSTITUTIONALWEAKNESSES

There are a number of sources of potential weak-ness including a lack of technical capability, poorspatial and jurisdictional coordination, poor func-tional coordination, and poor operational coor-dination.

The weakness of urban transport institutions isfrequently noted in this review. Few cities haveany comprehensive strategic land-use and trans-portation-planning agency. There is often no traf-fic management institution or unit worthy of thename, and often not even an adequately trainedtransport or traffic engineer within the municipalinstitution. Traffic police are often involved withtraffic management planning simply becausethere is no professional civilian alternative. Theyare often underequipped, not well trained in traf-fic management enforcement, and do not appre-ciate the role and function of traffic management.

The public transport planning and regulatory func-tion is too often tied to operations. If the institu-tions exist, they tend to be understaffed and theirstaff poorly trained. The net result is that politi-cians and senior executives without transporta-tion training are making important transportdecisions without informed professional input andwith predictable adverse consequences.

If the professionalism of the relevant institutionscannot be improved, it is unlikely that the substan-tive problems will be solved. Action is thereforerequired on two levels. First, authorities must rec-ognize what kind of technical organization is nec-essary to address urban transport issues. Second,the organizations must have adequate human andphysical resources to perform their functions.

SPATIAL AND JURISDICTIONAL COORDINATION Suburbanization of population growth, particu-larly where associated with continuing central-

TABLE 11.1 ALLOCATION OF STRATEGIC FUNCTIONS

Strategy level Function Agency Comments

“For the city” National roads Ministry of construction Private sector constructionPublic enterprises Ministry of economy Sometimes municipal

Tax levels TreasuryIntergovernmental transfers TreasuryRegulation and competition Ministry of economy May be function of a quasi-

policy independent commissionVehicle registration Ministry of transport

and safety

“Of the city” Urban structure planning Planning departmentStrategic transport planning Transport department

Local road management Roads departmentPublic transport planning Public transport agency

and procurementTraffic management Traffic department Direct responsibility to mayor

Law enforcement Police department Sometimes nationalRoad safety Interdepartmental unit

“In the city” Public transport operations Private companies Franchised or contractedRoad construction and Private companies Some force account

maintenance maintenance typicalLocal facility consultation NGOs and individuals Sometimes under formal public

inquiry laws

Note: NGO = nongovernmental organization.Source: Authors.

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 155

ization of employment and economic activity,leads to a situation where the major traffic flowsare between traditional local government juris-dictions rather than within them. The city region,rather than the municipality, becomes the effec-tive unit of major transport interaction.

Overlapping levels of authority within a hierar-chical system are often the source of jurisdictionalconflicts. In many transitional economies, respon-sibility for financing urban public transport has beentransferred to the municipalities, while the centralgovernment retains some powers (often embodiedin law) to determine fare levels. In countries with afederal structure, it is common to find very com-plex allocations of responsibility for urban trans-port between the central government, stategovernment, and municipality. In Brazil, for exam-ple, suburban railways were, until very recently, theresponsibility of the federal government, inter-municipal buses the responsibility of the state government, and intramunicipal buses the respon-sibility of the municipalities. In many cases the majormetropolitan areas are made up of severalautonomous municipalities. Particularly where polit-ical control is in the hands of different political par-ties, agreement may be very difficult to reach.

Contiguous authorities, at the same hierarchicallevel, make up many metropolitan areas in devel-oping, as well as in industrialized, countries. Wheretransport interaction occurs significantly across theboundaries between them, it is desirable that theyshould act in concert. Sometimes there is a domi-nant central authority whose leadership is accept-able to all others. That is particularly likely to be thecase where, as in Bangkok, the capital city has aspecial legal status or is more favorably treated bythe central government than are other local author-ities. More commonly, however—as in Manila,Caracas, and Lima—multiple jurisdictions of equiv-alent size and status may be unwilling to yield anysignificant amount of power or financial control toanother authority at the same hierarchical level.

Local district and municipal interests may notbe well aligned on detailed matters of transport

policy even in unitary cities. Local impacts of newinfrastructure and of public transport service andfare-level changes may be much better under-stood at the neighborhood level than at the citylevel. For this reason, even countries with anactive local democratic process, commonly sub-ject major transport infrastructure to a parallelprocess of detailed local consultation and review.

FUNCTIONAL COORDINATIONThe inability to avoid congestion by simplyincreasing road capacity leads to interactionbetween modes of transport (particularly betweenprivate cars and public transport) and betweenfunctions (particularly between land-use man-agement, infrastructure provision, road-trafficmanagement, and public transport planning andoperation). The transport system, rather than theindividual mode or function, becomes the effec-tive unit of transport supply.

Coordination between land-use and transportdevelopment is fundamental to efficient city devel-opment. Singapore and Curitiba, the two devel-oping cities widely considered to have bestmanaged their motorization, benefited immenselyfrom the early establishment of a long-term struc-ture plan. In both cases the structure plans havepermitted rational procurement of land for futurerights-of-way for main transport arteries and werethe basis for the location of high trip generators(such as shopping centers) in locations convenientlyaccessible by public transport. It has been possi-ble to provide a high level of accessibility to majoractivities by public transport and, as a consequence,to reduce the dependence on the private car. Inboth cities the institutional basis for this success-ful structure planning experience has been theearly establishment of a strong technical planningcapability.

Coordination between transport modes isrequired to secure effective physical interchangebetween modes, particularly when the modesare independently and privately operated. It isalso necessary to ensure efficient economicexploitation of the potential of all modes within

156 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

a system while at the same time maintainingaffordable basic service for the poorest. As far aspublic transport is concerned, cities that haveinstitutionalized a central functional responsibil-ity for planning and procurement appear to bemost capable of achieving both technical andeconomic coordination of the modes.

Coordination in traffic management and enforce-ment depends critically on the relationshipbetween transport policy and policing functions.In most countries police have extensive, and oftenexclusive, powers to direct road traffic, supportedby powers of arrest. This gives them a critical rolein policy implementation in traffic management,road safety, public transport priority, nonmotor-ized transport (NMT), the informal transportsector, and so on.

OPERATIONAL COORDINATIONPrivate and public sectors interact increasinglyas central and municipal governments find them-selves incapable of financing their strategies alone.Unfortunately, their objectives do not naturallycoincide. What is external to the private enter-prise is internal and central to the enlightenedmunicipal government. Incompatible and unco-ordinated development of infrastructure (as inBangkok’s expressways and mass rapid transit sys-tems) creates long-lasting burdens on the trav-eler. Unregulated operational competition createsenvironmental and safety hazards. This does notmean that the private sector cannot contributeimportantly to the resolution of urban transportproblems. But it does mean that there must beappropriate public institutions for the planning,procurement, and regulation of private sectorservices in order to reconcile the different interestsand efficiently mobilize private participation.

ORGANIZATIONAL OPTIONS

These coordination issues are difficult to handlebecause transport is but one of a number of serv-ices with differing degrees of spatial interactionand hence different “natural spatial boundaries.”

Even within transport, some functions, such asparking enforcement, may be appropriatelyadministered at a very local level, while others,such as infrastructure planning, need to beaddressed on a much wider scale. Moreover,countries have very widely differing political sys-tems and administrative competencies. There isthus no unique, universally applicable, structuralsolution. Arrangements may even change overtime within a country as priorities, problems, andthe political attitude toward planning change. Forexample, the United Kingdom has fluctuatedbetween joint transport committees of smallerautonomous authorities and allocation of primarypowers to city-region jurisdictions. Nevertheless,some general guidelines may be derived fromexperience in countries that have addressed theproblems reasonably successfully.

ESTABLISHING THE NECESSARY FUNCTIONAL BASIS The basic organizational requirements for goodurban transport are that each major function isrecognized, that responsibility for each functionis clearly assigned to an identified managementunit, that the units are properly resourced for theirtasks, and that their relationship with other organ-izations is clearly designated. The majority of localplanning and day-to-day operational functionswill fall to the municipal or metropolitan level. Atypical organization for the performance of thesefunctions, together with the responsibilities andresource requirements of the departments oragencies, is shown in table 11.2.

Managing public transport is particularly difficultwhen there are many modes, operating overboundaries of autonomous municipalities, somein the private and some in the public sector. Inthese circumstances, it is important that respon-sibility to ensure the coordination of physical infra-structure, service systems, fares, and finances isclearly allocated.1 There are several different pat-terns for this, including:

• A regional coordinating committee com-posed of political representatives of all the

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 157

jurisdictions, but without its own executivepowers. This sort of institution has been setup in several Brazilian metropolitan areas.

• A regional coordinating authority governedby a board of political representatives of theconstitutent authorities, having a profes-sional executive agency to implement itspolicies, with operators either directly con-trolled by the executive or operating undercontract to it. This was the pattern of thePassenger Transport Authority/Executivearrangement in the United Kingdom untilthe mid 1980s. Some of the German“verkehrversbund” (for example, Hamburgand Karslruhe) also follow this pattern.

• A regional mixed coordinating authoritygoverned by a board containing both polit-ical representatives of the constituent author-ities, and operators. This is the form of“verkehrversbund” in Stuttgart.

• A two-tier arrangement, in which there is apolitical body and an operators’ body linkedby a formal agreement. This is the arrange-ment in Berlin.

• A legally established independent author-ity governed by a broadly based representa-tive board of directors, including directorsnominated by the political jurisdictions but notunder direct political control. An example of thisstructure is the Madrid Transport Consortium.

The distinguishing feature of good examples ofeach of these structures is the ability to managethe system as one, and to introduce and main-tain integrated service planning and pricing. Forexample, the recovery of public transport patron-age in Madrid is closely linked to the establish-ment of the Madrid Transport Consortium andits introduction of a multimode transit “commutercard” (figure 11.1).

An appropriate starting point for a city wishingto improve its urban transport capability mightusefully be to fill out and expand a basic func-tions table and, by benchmarking on the organ-ization and capabilities of peer cities, to identifywhere its institutional and human resource weak-nesses lie (table 11.2).

FIGURE 11.1 EFFECT OF TRANSPORT INTEGRATION IN MADRID

1029.2

1141.9

1183.8

1014.5

1100

1149.8

1280.1

1327.6

1364.8

1236.31210.1

950.5

1043.5

900

1,000

1,100

1,200

1,300

1,400

1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

TOTA

L TR

IPS

IN P

UB

LIC

TR

AN

SPO

RT

(IN

MIL

LIO

NS)

CONSORTIUMINCEPTION

COMMUTER CARDESTABLISHMENT

Source: Madrid Transport Consortium 1999.

158 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

TAB

LE 1

1.2

P

RO

FE

SS

ION

AL

OR

GA

NIZ

AT

ION

FO

R M

UN

ICIP

AL

TR

AN

SP

OR

T F

UN

CT

ION

S

Rel

atio

nshi

p

Pri

ncip

al

Po

licy

Pro

fess

iona

l to

oth

er

Ref

eren

ces

Func

tio

nre

spo

nsib

iliti

esfu

ncti

ons

skill

so

rgan

izat

ions

Co

mm

ents

in r

evie

w

Urb

an s

truc

ture

Pr

epar

e an

d m

aint

ain

Shap

e d

evel

op

men

t La

nd-u

se p

lann

ers,

R

esp

ons

ible

to

may

or

Cha

pte

r 2

pla

nnin

gm

etro

po

litan

st

ruct

ure,

cre

ate

bas

is

envi

ronm

enta

l o

r ci

ty c

oun

cil

stru

ctur

e p

lan.

for

dev

elo

pm

ent

spec

ialis

ts,

cont

rols

.so

cio

log

ists

Stra

teg

ic t

rans

po

rt

Co

nduc

t st

rate

gic

Pr

epar

e b

road

Tr

ansp

ort

pla

nner

s,

Res

po

nsib

le t

o

Mus

t b

e co

nsis

tent

with

C

hap

ters

2, 3

, p

lann

ing

tran

spo

rtat

ion

stud

ies

stra

teg

ies

that

ec

ono

mis

ts, c

ivil

stra

teg

ic t

rans

po

rt

the

stra

teg

ic la

nd-u

se

9, 1

0Pr

epar

e co

mp

rehe

nsiv

eo

ther

org

aniz

atio

ns

eng

inee

rsau

tho

rity.

pla

nnin

g a

nd fu

nctio

n.tr

ansp

ort

atio

n p

lans

sho

uld

follo

w.

Rec

eive

s in

put

fro

m

for

the

city

or

oth

er m

unic

ipal

m

etro

po

litan

are

a.

tran

spo

rt u

nits

in

pre

par

ing

str

ateg

ies

and

pla

ns

Traf

fic m

anag

emen

tPr

epar

e tr

affic

D

eter

min

e tr

affic

Tr

affic

eng

inee

rs,

Res

po

nsib

le t

o

Som

e ac

tiviti

es (s

uch

Cha

pte

rs 2

, 3,

man

agem

ent

pla

ns;

prio

ritie

s co

nsis

tent

ec

ono

mis

ts,

stra

teg

ic t

rans

po

rt

as in

spec

tion

and

4,

6, 9

revi

ew d

evel

op

men

t w

ith g

ener

al

par

king

sp

ecia

lists

, au

tho

rity.

mai

nten

ance

) can

be

pro

po

sals

with

tra

ffic

stra

teg

y.el

ectr

ical

M

ust

wo

rk in

co

ntra

cted

out

to

im

pac

ts; o

per

ate

Cre

ate

par

king

and

en

gin

eers

coo

rdin

atio

n an

d

priv

ate

sect

or.

Traf

fic C

ont

rol a

nd

traf

fic m

anag

emen

t co

nsul

tatio

n w

ith

ITS

(Inte

llig

ent

Tran

spo

rtfr

amew

ork

for

loca

l po

lice

Syst

ems)

; man

age

low

er-le

vel d

istr

icts

.au

tho

rity.

vehi

cle

insp

ectio

n an

d m

aint

enan

ce

sche

me;

mo

nito

r en

viro

nmen

tal i

mp

acts

.

Pub

lic p

asse

nger

Pl

an a

nd r

egul

ate

Prep

are

pas

seng

er

Pub

lic t

rans

po

rt

Res

po

nsib

le t

o

Pro

cure

men

t p

refe

rab

ly

Cha

pte

rs 3

, 7tr

ansp

ort

pub

lic t

rans

po

rt

tran

spo

rt p

olic

ies

and

reg

ulat

ory

st

rate

gic

tra

nsp

ort

b

y an

arm

s le

ngth

sy

stem

s in

clud

ing

co

nsis

tent

with

sp

ecia

lists

auth

orit

y.ag

ency

.b

uses

, tra

ms

and

st

rate

gy

and

with

Sh

oul

d b

e se

par

ate

light

rai

l, ta

xis,

and

fin

anci

al c

apab

ility

. fr

om

any

pas

sen-

met

ros.

Se

t p

aram

eter

s fo

r g

er t

rans

po

rt

Pro

cure

ser

vice

s.p

rocu

rem

ent

agen

cy.

op

erat

ions

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 159

Traf

fic e

nfo

rcem

ent

Enf

orc

e tr

affic

C

olla

bo

rate

in t

raffi

c Po

lice

offi

cers

Traf

fic p

olic

e p

rovi

de

Traf

fic p

olic

e ty

pic

ally

C

hap

ter

6re

gul

atio

ns; m

anag

e m

anag

emen

ttr

affic

acc

iden

t an

d

par

t o

f a c

ity p

olic

e tr

affic

eve

nts

and

sy

stem

des

ign.

traf

fic in

cid

ence

d

epar

tmen

t w

ith w

ider

in

cid

ents

; co

llect

E

nfo

rce

traf

fic

info

rmat

ion

to s

afet

y en

forc

emen

t fu

nctio

ns.

acci

den

t d

ata.

man

agem

ent

and

tra

ffic

man

agem

ent

po

licy.

o

rgan

izat

ion.

Ro

ad d

esig

n,R

esp

ons

ible

for

Mai

nten

ance

C

ivil

eng

inee

rsW

ork

clo

sely

with

Tra

ffic

Co

uld

be

par

t o

f a c

ityC

hap

ters

3, 6

, 9co

nstr

uctio

n,

des

igni

ng, c

ons

truc

ting

, p

riorit

izat

ion

Man

agem

ent

pub

lic w

ork

s an

d m

aint

enan

cean

d m

aint

aini

ng s

tree

ts.

Org

aniz

atio

n to

d

epar

tmen

t.im

ple

men

t d

etai

led

tr

affic

eng

inee

ring

wo

rks

Traf

fic s

afet

yR

oad

tra

ffic

safe

ty

Ana

lyze

saf

ety

dat

a;

Stat

istic

ians

; D

irect

ly r

esp

ons

ible

M

ust

be

a p

rimar

y C

hap

ters

5, 9

stra

teg

y. C

oo

rdin

ate

orc

hest

rate

inte

r-tr

affic

eng

inee

rsto

may

or

or

city

re

spo

nsib

ility

of a

al

l dep

artm

enta

l d

epar

tmen

tal

coun

cil.

Rel

atio

nshi

p

lead

dep

artm

ent,

in

put

s, in

clud

ing

co

llab

ora

tion

to

with

hea

lth

such

as

traf

fic

tho

se fr

om

hea

lth,

imp

lem

ent

stra

teg

y.au

tho

ritie

s ne

cess

ary

eng

inee

ring

. Po

lice

educ

atio

n, a

nd s

o o

n.co

mm

itmen

t es

sent

ial.

Sour

ce: A

utho

rs.

160 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

REDUCING THE SCOPE FOR HIERARCHICAL CONFLICT Interactions between different layers of govern-ment may be very complex. Very clear andexplicit separation of function between levels inthe hierarchy may avoid many problems. Forexample, in the process of decentralization inBrazil, the federal government transferredresponsibility for suburban rail operations to thestates. That has certainly contributed to improv-ing modal coordination in major Brazilian cities.Even with strategic issues retained at the met-ropolitan or regional level, however, the inter-ests of individual lower-level jurisdictions need tobe taken into account.

It does not always work, however. In the Braziliancase, transfer was usually smooth because thefederal government accompanied the transferwith appropriate financial arrangements, gener-ally consisting of a thorough rehabilitation of thecapital stock before transfer. In contrast, in manyof the newly independent countries of the formerSoviet Union, the transfer of responsibility wasnot accompanied by a transfer of funding. Insome cases the responsibility of funding publictransport was even transferred to municipalitiesor regions, while the central government retainedpowers to determine fares or to grant discountsor exemptions. The result has usually been a dis-astrous decline in service.

COORDINATING OVERLAPPINGAUTHORITIESThere are a number of different ways in whichcoordination between authorities may be sought.At one extreme is the creation of a single-tierconurbation authority with plenary powers, notonly for transport but also for associated func-tions, such as land-use planning and control. Aless monolithic arrangement is to have a two-tierlocal government structure, with powers for trans-portation planning and conurbation transportpolicy formulation assigned to the conurbation-wide level and other, less strategic, responsibili-ties, such as local parking control, left for exerciseat the second level. In these arrangements the

high-level authority has its own independentbudget with either its own direct revenue sourcesor predetermined allocations from either loweror higher levels of government. A third, less rad-ical, arrangement would be the creation of a jointcommittee of the independent authorities. Theparticipants on the committee would remainfinancially autonomous and agree on the alloca-tion of costs among them.

The difficulty with associations based on voluntarycollaboration is that there will be significant like-lihood of rent-seeking behavior in the dealingsbetween authorities. Three institutional devicesreduce that possibility.

First, there should be a national legal frameworkwithin which collaboration between authoritiesmust take place. This is clearly established by lawfor public transport in France.2

Second, the authorities should have some free-dom to determine the boundaries of the collab-oration. In France this is done by authoritieshaving the freedom to set up their own planningboundary by agreement.

Third, there may need to be some externallyinjected financial inducement to collaborate. InFrance, this is achieved by requiring the estab-lishment of an organizing authority as a condi-tion for being able to levy a local transport tax(the “versement transport”) or to receive certaincentral government contributions. In Brazil,development of regional transport coordinatingcommittees was a condition of financing urbanrail developments in the major cities. Thisarrangement may be strengthened by making allexpenditure involving government aid subject tocertification by the regional authority that it is inaccordance with policies for the region.

The metropolitan authority would typically becharged with:

• Integrating strategic urban land-use andinfrastructure planning with transport system

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 161

and network planning, including the devel-opment and publication of a strategic plan-ning framework for transport and land usein the metropolitan region

• Integrating road network planning withpublic transport planning

• Integrating the planning, policy, regulation,and pricing of the various public transportmodes.

COORDINATING FUNCTIONS—PLANNING AND OPERATIONSRecognition of the interaction between modesand functions requires the development of mutu-ally compatible policies through a process ofstrategic transport planning. Often this is embod-ied in a transport and land-use plan, or structureplan, which offers the strategic framework within

which separate but consistent functional andmodal policies are developed. Shorter-term plansoften take the form of a rolling five-year programand an annual plan and program. Modern man-agement tools, such as planning, programming,and budgeting systems, and the adoption of arange of standards and guidelines for service pro-vision can assist in converting good planning intowell-managed implementation. A typical logicalstructure for the planning process is shown infigure 11.2.

A wide range of strategic functions is usuallydevolved to the metropolitan region level agency,including:

• Land-use development strategy• Environmental strategy

FIGURE 11.2 A TYPICAL METROPOLITAN PLANNING PROCESS

STRATEGICINPUTS

PLANNINGDOCUMENTS

20-YEAR STRUCTURAL PLAN,

5-YEAR ROLLING PLAN, AND

ANNUAL PLAN AND PROGRAM

ADMINISTRATIVEORGANIZATION

NATIONALPLAN ANDBUDGET

LOCALPLAN ANDAGENDA

PUBLICTRANSPORTOPERATIONS

TECHNICALREGULATION

ANDENFORCEMENT

ECONOMICREGULATION

ANDPROCUREMENT

DIRECTWORK

EXECUTION

MUNICIPALPLANNING,

ORGANIZATIONPLAN, PROGRAMS,

AND BUDGET

Source: Authors.

162 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

• Road planning, including supervision of pri-vate concession development

• Traffic management strategy• Parking and road pricing strategy• Public transport planning and service pro-

curement.

Assigning this range of functions to the metro-politan region authority requires it to developintegrated strategies, knowing that, subject tothe limits of a normal democratic process, it hasthe ultimate power to implement them. Theimplementing agencies should operate withinthis overall strategic framework, whether they aresubsidiaries to the metropolitan region author-ity, quasi-independent agencies, or units withinexisting ministries. Where enforcement is an inte-gral part of a function, the resource requirementsfor enforcement should be recognized and allo-cated as part of the decision process. The criti-cal issue is that the agreed-on strategy fortransport and land-use planning should takeprecedence over ad hoc, independent initiatives.

INTEGRATING IMPLEMENTATION AND ENFORCEMENT—THE ROLE OF THE POLICETraffic management faces one very intractableinstitutional element in most countries. Whiletraff ic management measures should bedesigned to be as self-enforcing as possible(physical dividers on bus lanes, guardrails to pre-vent illegal parking, and so on), some policeenforcement is always likely to be necessary.While in some cities traffic police have provedto be the only agency ready to take action toresolve serious traffic issues (such as the instal-lation of median strips in Colombo to preventtraffic accidents at crossings), they tend to oper-ate to their own agenda. Their main objective isusually to “keep traffic moving,” particularly onmain routes. For example, the failure of area traf-fic control in Bangkok is largely attributed to thepolice propensity to override the system.Similarly, successful enforcement of with-flowbus lanes, standard in industrialized cities, hasproved difficult in many developing cities. The

inability to develop a strong traffic safety cam-paign in Buenos Aires was similarly attributed todifficulties in coordination between police andtransport agencies.

That independence of action has institutionalroots. Traffic police are often organized at anational level, responsible to a central ministryof the interior, and are often suspected of beingboth corrupt and politically partisan. In somecountries, such as Mexico and Venezuela, citytraffic police forces also exist and work in paral-lel with the national force. In some countries,traffic police are separated from the generalpolice force. Even where they are, as in theRussian Federation, the law allows the trafficpolice to be used for the maintenance of publicorder. A poor police image inevitably damagesthe credibility of, and respect for, any traffic meas-ures—such as bus lanes and speed control—thatrely on strict and impartial enforcement. It alsocreates a reputational risk for multinational donoror lending agencies, such as the World Bank,with a responsibility under charter to avoidinvolvement in local political matters.

Despite these problems it is difficult to makeprogress in urban transport management with-out the involvement of the police. The WorldBank has therefore financed transport projectswith traffic police components in all regions ofthe world.3 Police components usually compriseequipment to assist traffic regulation enforce-ment, technical assistance for traffic accidentreporting and analysis systems, and technicalassistance for traffic regulation enforcement.Efforts to mitigate the risk of misuse of invest-ments include such measures as ensuring thatequipment financed was not of a kind that couldbe used for nontraffic purposes, monitoring ofequipment use, and covenants on traffic policeactivities.

The recommended institutional basis for thisinvolvement is that the traffic authority should beresponsible for planning, design, and operationof traffic schemes, including computer-controlled

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 163

traffic signal systems.4 Traffic police responsibil-ities should be confined to traffic regulationenforcement, but they should be involvedthrough consultation in scheme design and ona continuous basis in an advice and assistancerole. Securing such coordination and coopera-tion is likely to be more easily achieved wherepolice and transport authorities are responsibleto the same jurisdiction.

COORDINATING PUBLIC AND PRIVATESECTORSThere is no need for a strategic planning author-ity to have any direct operating responsibility orto own any transport service supply capacity.Success in mobilizing private sector participationin urban transport has already occurred in anumber of areas, including:

• Investment in new road and public transportinfrastructure

• Concession of existing public transport oper-ations, with responsibility for future capitalfinance

• Concession or privatization—whole or inpart—of ports and airports.

In all of these areas, institutional arrangementshave been necessary to ensure that there is abasis for commercial operation that is consistentwith an integrated urban transport policy.

The first step in a successful involvement of theprivate sector in urban transport is the separa-tion of planning from operations to ensure thatthe benefits of competition can be obtained with-out the strategic planning agency’s feeling theneed to protect its “own” operating agency.Restructuring of existing line supply agencies isusually necessary to achieve this, because nationalor municipal transport undertakings typically per-form both planning and operational functions.That is not an insuperable problem. In severalindustrialized countries, the planning and pro-curement skills of the monopoly agency havebeen transferred to a quasi-independent pro-curement agency. However, as the nature of the

procurement task is changed from the procure-ment of goods to the procurement of transportservices, some additional skills may need to bedeveloped.

The residual operating function is also likely toneed restructuring. The fact that the public sectorcurrently has a monopoly of the necessary skillsis often used as an argument that competition isimpossible. In many countries the central gov-ernment has washed its hands of formal financialresponsibility for its traditional supply agenciesby converting them into state joint stock com-panies, the majority of shares of which may still beheld by a state property agency. Experience sug-gests that unless a much more thorough sepa-ration is made, and a real opportunity forindependent management and initiative created,the change will have little effect. In reality noth-ing is immutable. The operating company canbe restructured as a number of independentprofit centers to compete with each other andeventually to be privatized. This kind of transfor-mation may need some time (it took nearly 10years to complete in London), but can be doneif the political will is there.

Capacity building in the private sector is often akey issue and should be taken into account indesigning policies and programs. Where somecompetitive potential already exists in a frag-mented, small-scale private sector, it needs someinstitutional fostering. This may take two forms.First, it may be necessary to create a legal basisfor the establishment of operators’ associationswith the appropriate standing to engage in com-petitive tendering competitions. Such associa-tions were long the basis of public transportoperations in Buenos Aires, and have morerecently been provided for in law in some of theCentral Asian countries, such as Uzbekistan,where tendering is being introduced. Second, amore difficult point is that attention may need tobe given to the institutional arrangements forvehicle finance, and particularly the legalarrangements for vehicle leasing. The weaknessof these institutions in some of the transitional

164 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

economies has been a serious impediment tothe transformation of the private sector from alow-quality, peripheral, service provider to a main-stream part of the public transport system.

Restructuring of government budget arrange-ments may also be necessary. In many countries,and particularly in the transitional economies ofEastern Europe, public transport operators arelegally obliged to offer free or reduced-fare trans-port for many categories of passengers. Theadverse impacts of these obligations on the sus-tainability of service and on the welfare of thosewho most need support have been discussed inchapter 3. To avoid such impacts, and to ensurethat resources devoted to the social services areused to their best effect, it is desirable that alltransport subsidies supporting other sector objec-tives (health, education, police, and so on) shouldbe directly funded through those sector budg-ets. Where the finances are not available to fundthe exemptions, operators should be under noobligation to honor them.

Some regulatory requirements will remain. As faras technical conditions of vehicle constructionand use are concerned, it is likely that these willremain the function of a national regulatoryagency, both for rail- and road-service providers.General business behavior will need to be con-trolled by whatever general regulator is respon-sible for control of monopolization or of restrictiveand predatory practice (although this is often weakin developing countries). Where there has alreadybeen a broad extension of private provision oftraditional public utility services, a regulatoryagency responsible for adjudicating priceincreases under concession contracts may alreadyexist. Where concession contracts themselvesattempt to cover the conditions of operation ofthe concessions completely, it may be necessaryto ensure that the relevant courts have the skills forthe kind of contract adjudication that might arise.

ENGAGING CIVIL SOCIETYEven in industrialized countries with well-devel-oped, formal, local democratic processes, it is

common practice for civil society to be engagedin validating major infrastructure investments orpolicy changes. This may occur through advanceexposure of plans to a free press and other media,as well as through more formal processes ofpublic consultation or public inquiry. This partic-ipation can operate at several levels. For small-scale localized projects, it may be possible toincorporate local preferences in the designprocess itself. For example, local residents’ asso-ciations in Lima have been involved in identify-ing desired improvements in access roads toinformal housing areas. Similarly, public transportusers may be involved in service franchisingarrangements through provision for direct com-plaints to the regulating authority (as in BuenosAires), for consumer consultation on new routes,and through the involvement of the media inassessing performance of franchised operators(as in Bogotá).

At a more strategic level, and for larger, morecomplex, projects, consultation often functionsmore as a means of giving voice to local concernsin the process of trying to reconcile inherentlycompeting or conflicting interests. It is inherentlymore difficult to mobilize citizen involvement atthe strategic level in the development of con-sensus-based city development strategies.

Developing such strategic involvement requiresaction at two levels. First, particularly where formallocal political processes are weak, the existenceof effective local community groups is extremelyimportant. In developing countries such groupsare often well developed in rural areas, but muchless so in cities. The Bank can strengthen thisdevelopment by very positively seeking theinvolvement of urban community groups in proj-ect implementation. There are some promisingexamples. In Mumbai, slumdwellers’ organiza-tions have been involved from the outset in plan-ning and implementing resettlement associatedwith improvement in the urban railways (box 11.1).Efforts are also being made to encourage a moreinclusive decisionmaking process in Indonesia.To aid this process, the World Bank has devel-

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 165

oped a participation toolkit, bringing togetheradvice on good practice in the organization ofparticipation.

Second, the public processes must be organizedto facilitate timely but well-informed consulta-tion. An interesting example of this is the pre-sentation of long-term transport strategies toban cars in the center of Bogotá during peakhours after the year 2015 in a referendum inBogotá in September 2000. While such long-termcommitments are susceptible to being over-turned in the future, the significance is to raisethe level of interest, understanding, and debateon critical strategic issues.

CAPACITY BUILDING

Skills development is required in a number of dif-ferent areas, including planning and engineer-ing, and traffic management and enforcement.

Other important issues that need to be addressedinclude how to distribute and retain key staff, andwhen to train them.

PLANNING AND ENGINEERINGInstitutions responsible for urban transport indeveloping countries, whether at a local ornational level, are generally short of adequateprofessional skills. While there may be enoughhighway or construction engineers, other pro-fessional staff (such as traffic engineers, transportplanners, economists, and public transport reg-ulators) tend to be in short supply. In Bangkokand Jakarta, it is conservatively estimated thatthere are about one-tenth the number of trafficengineers that would be found in cities of a sim-ilar size in industrialized countries.

Technical and administrative staff to support theprofessional staff may be adequate in number,but also frequently lack necessary skills. Eventhose staff who do have technical skills in urban

BOX 11.1 ENGAGING CIVIL SOCIETY IN PLANNING RESETTLEMENT OF THE POOR IN

MUMBAI

The Mumbai Urban Transport Project envisages improvements in urban transport in Mumbai, pri-marily to encourage public transport. The project will require resettlement of over 75,000 people,most of whom are squatters. Because of the size of resettlement, project preparation has empha-sized a strong participatory approach and involvement of Mumbai’s civil society in the project.This is being achieved through the reliance on local nongovernmental organizations (NGOs) aspartners. The NGOs are responsible for community organization and data gathering, and con-struction of temporary and some permanent housing. This process is being carried out in a par-ticipatory manner with groups of project-affected people. The project resettlement policy wasalso prepared in a participatory manner using a task force headed by a retired chief secretary withmembers from government agencies, local NGOs, and the private sector. The project will alsoestablish an independent monitoring panel of eminent citizens of Mumbai who will ensure thatthe Bank’s safeguard policies are followed, and provide a last-resort grievance redressal mecha-nism for the affected people. A main problem with the participatory process has been reconcil-ing the use of NGOs as partners with the Bank’s procurement guidelines. There has been closecollaboration among all parties to find solutions that would meet both the Bank’s procurementguidelines and the objectives of partnership with local NGOs.

Source: Harald Hansen, from World Bank project files.

166 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

transport may not be in positions to make themost effective use of them, having achievedmanagement status outside their technical fieldor being in units where administrative, jurisdic-tional, or functional limitations restrict the extentto which they can apply their specialized knowl-edge. Many trained staff lack the necessary expe-rience to guide them through the social andpolitical issues they face in applying their tech-nical expertise. Technical support tools such aslocal design standards and guidelines, goodpractice manuals, and more sophisticated knowl-edge management systems also tend to beunderdeveloped.

TRAFFIC MANAGEMENT ANDENFORCEMENTTraffic management skills are particularly scarce.Although traffic management schemes are ofrelatively low cost, they require a high staff inputand the resolution of many detailed and inter-related planning, design, and procurementissues. Success depends on staff capabilities andthus on staff training. Countries such as Braziland Chile have a well-established traffic man-agement capability and staff training, but in manycountries traffic management is not recognizedas a distinct discipline, and staff capabilities andtraining procedures are deficient. Capacity build-ing can be assisted in a number of ways. Trafficmanagement can be introduced as an estab-lished part of curriculums in institutes of highereducation. Some universities in industrializedcountries already have “twinning” arrangementswith universities in developing countries, oftenwith the assistance of bilateral funds. There havealso been successful examples of twinning oftransport and traffic agencies as betweenGothenburg (Sweden) and both Hanoi (Vietnam)and Ibadan (Nigeria).

Traffic police training is mostly carried out atpolice-run training centers, and normally empha-sizes routine policing (such as license checking,stolen vehicles, city-checkpoint security controls,and so on) rather than traffic management mat-ters (traffic scheme enforcement, particularly

selective enforcement techniques, gaining famil-iarity with traffic laws and regulations, dealingwith traffic emergencies, accident reporting andanalysis, and so on). As with training for profes-sional transport planners and engineers, a sys-tematic approach to traffic police training isneeded. Training for traffic police is a specialisttask: few traffic engineers have an adequate back-ground to identify and plan such programs.Programs can only be devised by traffic policespecialists. Furthermore, operational methods oftraffic police forces vary widely from country tocountry, and training must conform to local cul-tures (although it will sometimes be advanta-geous to improve operational practices) and havea high local input in its preparation.

DISTRIBUTION AND RETENTION OF SKILLSWhat professionals there are tend to be con-centrated in central government departments orin metropolitan areas, which are not necessarilywhere they are most needed. Making the bestuse of scarce professional skills may call for someinstitutional flexibility. Where problems are great-est in a capital city, it may be sensible to establisha specialist unit in the capital city, which isdesigned to develop people and techniques thatcan be used elsewhere in the country. For exam-ple, the London Traffic Management Unit, thoughit had a London remit, boosted traffic manage-ment capability throughout the United Kingdom.In developing countries a similar approach isbeing attempted; an example is the Office of theCommission for Management of Land Transportin Bangkok. Such units need to be linked tonational research and development agencies, aswell as to leading academic institutions. Supportfor such breeding grounds for technical skills maybe an essential instrument for the multilateraldevelopment agencies.

Not all the skills need to be in the public sector.Particularly where government salary levels makeit difficult to retain staff, it may be more efficientand cost-effective for scarce specialist skills tobe maintained in centralized specialist technical

STRENGTHENING URBAN TRANSPORT INSTITUTIONS 167

units and brought in when required by munici-palities. Another option is to focus on develop-ing the local consult ing industry and forgovernment agencies to contract out servicesto consultants on a project-by-project basis orcomprehensively on a term basis (as is beingdone in the United Kingdom). Whichever optionis chosen, the best results are likely to beachieved by spreading resources geographicallywhile developing a number of units as “centersof excellence,” particularly in more advancedfields such as transport modeling and IntelligentTransport Systems.

TRAINING INSTITUTIONSDeveloping the necessary human resources forurban transport is not a simple or short-term task.For basic training, particularly for technicians orcategories of staff where there is a high turnover,either in-house or external training may be ade-quate. Increasing the numbers of professionalspecialists requires an expansion of tertiary edu-cation coupled with subsequent practical pro-fessional training. The main options available arefor public sector transport organizations toexpand their own facilities, to expand the localtertiary education sector, or to use already exist-ing facilities in industrialized countries. In prac-tice a combination of all three options is likely tobe necessary in the short term. Over time theobjective should be to develop local technicaland tertiary education, with the public servicesectors gradually devolving training to these facil-ities. In parallel, a professional accreditation andlicensing system should be put in place. This canbe managed by the government and fostered bya learned society (as in the United States) or beboth fostered and managed by the learned soci-ety (as in the United Kingdom). Twinning withorganizations having similar technical responsi-bilities in developing and industrialized countriesenables staff to gain valuable practical technicaland city management experience. Regional spe-cial initiatives (such as the Latin America CleanAir Initiative and the Asia Clean Air Initiative)permit knowledge sharing on emerging topicsof common interest.

CONCLUSIONS: A STRATEGY FORINSTITUTIONAL REFORM INURBAN TRANSPORT

No single institutional blueprint for urban trans-port is appropriate for all countries. Nevertheless,there is enough experience of the difficulties aris-ing from the failure to align policies between juris-dictions and agencies, or to secure collaborationbetween them, to establish some general prin-ciples for the reduction of institutional impedi-ments to effective policy integration.

A strategy for institutional reform should addressthe main areas of human resource developmentand coordination, and be based on the follow-ing principles:

For functional capability and human resourcedevelopment• Municipal or metropolitan transport agen-

cies should establish an administrative struc-ture within which responsibility for allnecessary technical functions in urban trans-port are clearly identified and allocated.

• Central governments should develop a train-ing strategy for professional and technicalskills in urban transport.

• Scarce professional skills should initially beconcentrated, and retained by adequateremuneration, in either public sector “centersof excellence” or in private sector consult-ant organizations.

• Collaboration between authorities shouldbe encouraged both nationally, to shareavailable skills, and internationally, to furtherdevelop skills and experience.

For jurisdictional coordination• Allocation of responsibility among levels of

government should be clearly establishedby law.

• Intergovernmental transfers should be care-fully planned to be consistent with the allo-cation of responsibility.

• Formal institutional arrangements should bemade for collaboration where multiple

168 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

municipalities exist within a continuousconurbation.

• Central government should use intergov-ernmental transfer arrangements to encour-age coordination at the metropolitan level.

• Obligations statutorily imposed on localauthorities should be linked to specificchannels of finance (such as direct-lineagency funding of reduced-fare or freepublic transport).

For functional coordination• There should be a strategic land-use and

transport plan at the municipal or metro-politan level with which detailed planning,both of transport and land use, should bealigned.

• Functions should be clearly allocatedbetween agencies, with more strategic func-tions being retained at the higher level inmetropolitan areas.

• Traffic police should be trained in traffic man-agement and safety administration andshould be involved collaboratively in trans-port and safety policy planning.

• Responsibility for traffic safety should beexplicitly allocated with an institutionalresponsibility at the highest level of the localadministration (mayor’s office or its equiv-alent).

For effective involvement of the private sector• Planning and operating responsibility for

public transport should be institutionally sep-arated.

• Technical regulation should be separatedfrom procurement and economic regulation.

• Operating agencies should be fully com-mercialized or privatized.

• The development of new competitive privatesuppliers of service should be encouragedthrough legal recognition of associations, andso on.

• A clear legal framework should be estab-lished for competition in public transportsupply, either in the market or for the market.

• The public sector should develop profes-sional service procurement and contractenforcement skills.

NOTES

1. The alternative forms are discussed in detailin Prointec Inocsa Stereocarto 2000.

2. See Louis Berger S.A. 2000.3. The countries include Albania, Armenia,

Bangladesh, Brazil, China, Costa Rica, Egypt,Hungary, India, Lithuania, Pakistan, Poland,Romania, Turkey, and the República Bolivarianade Venezuela.

4. Barrett 1986.

DEVELOPMENT CHALLENGES

The general objectives of this review have been(a) to develop a better understanding of thenature and magnitude of urban transport prob-lems in developing and transitional economies,particularly with respect to the elimination ofpoverty; and (b) to articulate a strategy to assistnational and local governments to address urbantransport within which the role of the Bank andother agencies can be identified. This chaptercompletes the review by identifying how the Bankcan take forward the substantive agenda.

The context for the review has been continuingurbanization in developing countries and some,but not all, transitional economies, and increas-ing motorization of the urban areas. In that con-text, the focus has been on the “fundamentalparadox of urban transport,” namely, the coex-istence of excess demand for road space with aninability to adequately finance public transportservices. The review started by examining theurban transport–poverty nexus, arguing thaturban transport impacts on poverty indirectly,through its effect on economic growth (chapter2), and directly, through its impact on the lives of

the poor themselves. The direct impacts are notonly economic (chapter 3) but also relate to theenvironmental quality of life (chapter 4) and itssafety and security (chapter 5). These impactswere subsequently explored with respect to themain modes of private road transport (chapter6), public transport (chapters 7 and 8), and non-motorized transport (chapter 9). What emergedwas an explanation of the paradox in terms of acombination of inappropriate pricing and financ-ing mechanisms (chapter 10) and fragmentedinstitutions (chapter 11). In particular, the prob-lem is seen to result from the separation of roadcharging from road provision and finance, theseparation of the provision of urban transportinfrastructure from operations, and the inade-quate integration of modes within the urbantransport sector and of urban transport with therest of development strategy.

This diagnosis is neither new nor specific to thedeveloping world, but there are some emergingtrends, which pose new challenges, but also createnew opportunities for progressive reform. Themost notable of these trends are:

• Intensified focus on poverty reduction

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 169

MEETING THE DEVELOPMENT CHALLENGES:HOW CAN THE BANK CONTRIBUTE?12The urban transport scene is changing rapidly, with a sharper focus on poverty

reduction, decentralization of responsibilities, increase of private participation,

and deterioration of environmental, safety, and security conditions. These

developments pose new challenges, but at the same time create new

opportunities to redress the defective pricing and financing practices and

institutional fragmentation to which poor urban transport performance has been

attributed in the past. This final chapter suggests how the World Bank can best

support the development of sustainable urban transport institutions and policies

in these new conditions.

170 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

• Decentralization• Increasing private participation in urban

transport supply• Deteriorating safety and security• Increased concern about the local and global

environment.

In this chapter we present, for each of these chal-lenges, (a) a synthesis of the policies advocated,(b) a review of progress so far in assisting theimplementation of these policies, and (c) anagenda for World Bank action to take the poli-cies further.

STRENGTHENING THE FOCUS ONPOVERTY

The very poor often cannot afford public trans-port and are dependent exclusively on NMT,mainly walking. It is clear that few governmentsrecognize the importance of such modes to thepoor and devote inadequate attention andresources to them. In many middle-income coun-tries, the efficiency of public transport opera-tions is critical to the welfare of the poorer groupswho are primarily dependent on it. Governmentsfrequently impose general fare controls to ensurepublic transport affordability, fare controls thatare often not well targeted to the poor, that“leak” to suppliers or privileged user classes whoare not poor, and, where no external means areprovided to finance them, reduce the quality orquantity of service. The vulnerable poor areoffered insufficient protection. The main policyprescriptions to address these deficiencies are:

• Emphasize road and public transport accessto low-income housing areas.

• Give priority in road planning and manage-ment to the needs of public transport.

• Develop competitive regimes to reducecosts of public transport.

• Focus any fare subsidies narrowly, and fundthem securely.

• Protect against adverse spin-off effects, suchas resettlement and redundancy.

PROGRESS SO FAR The development of the Comprehensive Devel-opment Framework and Poverty ReductionStrategy Papers (PRSPs) has brought infrastruc-ture activity under increasingly critical scrutiny.While both governments and civil society con-tinue to stress the importance of infrastructure inmeeting the fundamental problems of exclu-sion,1 the emphasis in urban transport has clearlychanged. The emphases in Bank urban transportin the past decade include the following:

• Concentrating on the transport modes ofthe poor. In the current urban transport port-folio, more is being invested in public trans-port than in roads. Urban road maintenanceprojects, as in the Kyrgyz Republic, oftenconcentrate on the main public transportroutes. The Bank has encouraged the pro-vision of infrastructure facilities for NMT inAccra, Lima, and Manila; it has also—inLima—addressed the problems of financeof bicycles for the poor.

• Improving public transport efficiency. TheBank has long advocated competition as thebest way of ensuring public transport effi-ciency. However, experience now suggeststhat, particularly in the larger cities, compe-tition for tendered bus franchises within awell-monitored and enforced regulatoryframework may yield lower fares and bothenvironmentally cleaner and safer servicesthan total deregulation. It has supportedsuch reforms in Uzbekistan and Senegal. Inurban rail transport, Bank-supported con-cessioning to the private sector has alsoyielded significant efficiency improvementsin Argentina and Brazil.

• Encouraging sustainable public transportfares strategy. In projects in the RussianFederation and some of the Central Asianrepublics, the Bank has emphasized theimportance of sustainable financing throughcost reduction, direct funding of reducedfares by line agencies (health, education,and so on) and, where necessary to sustainservice, fare increases. Multimodally inte-

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 171

grated fare schemes to protect the poorhave been developed in Bank projects inseveral Brazilian cities.

• Protecting the vulnerable. During the pastdecade, the formal safeguard policies of theBank have been strengthened substantiallyto ensure that anyone whose residence orlocation of livelihood is directly affectedbecause of a Bank project is resettled orcompensated so that the change imposesno detriment to them. The Bank has alsofinanced redundancy and retraining com-pensation in rail reforms in Argentina, Brazil,and Poland both through project loan fundsand through adjustment lending.

THE AGENDA Despite these efforts the poverty focus is still notsharp enough. Better understanding is requiredof the impact of urban transport infrastructure andservice policies on poverty. There is a danger that,in the absence of effective demand management,road investments to eliminate bottlenecks will tendto benefit the rich at the expense of the poor. NMTand walking remain underfinanced, and the spe-cial needs of women and the mobility impairedare neglected. Competition is still seen too simplyas total absence of regulation, and hence is asso-ciated with undisciplined chaos. Formal require-ments for line agencies to finance subsidies totheir client groups are not honored. Resettlementand redundancy compensation is often viewed asa bureaucratic enclave requirement of Bank proj-ects, to be avoided by judicious choice of projectsfor Bank financing. The Bank can further assist instrengthening the poverty focus of its urban trans-port activities in the following ways:

• Emphasize the importance of facilities forNMT, particularly adequate sidewalks forpedestrians and creation of a friendlier reg-ulatory environment and better traffic man-agement for all nonmotorized modes.

• Seek to develop new forms of finance forbicycles for the poor.

• Require, as far as practical, the inclusion ofthe full economic, social, and environmen-

tal impacts, including those on affectedNMT users, in the formal cost-benefitappraisal of projects.

• Increasingly support the development of com-petitive franchising systems, through techni-cal assistance in the reform process, andwhere appropriate, through financing of refur-bishment of public transport systems in theprocess of reform. The technical assistancewill also assist countries in designing systemsthat, while permitting the informal sector toplay a continuing role, ensure that unfair com-petitive advantage is not taken at the expenseof safety or environmental quality.

• Encourage infrastructure and service designto better accommodate the mobilityimpaired.

• Emphasize the importance of institution-ally sustainable finance of public transportoperations. Where direct income transfersto the poor are not possible, do notoppose, in principle, well-targeted subsi-dies of services essential to the poor.

• Support general fare control only where thefinance necessary to maintain the systemis ensured or the implications of fare con-trols on supply have been carefully consid-ered and the quality of service implied bysuch controls explicitly chosen.

• Continue to emphasize adequate com-pensation for locational or occupational dis-turbance arising from projects.

• Encourage the further participation of non-governmental organizations in the designand implementation of adjustment arrange-ments and require evidence of early and ade-quate consultation among affected parties.

FACILITATING DECENTRALIZATION

In many countries (particularly in the former SovietUnion and in Latin America) responsibility forurban transport is being decentralized. Some-times this is merely a by-product of the failure oftraditional central fiscal sources. Frequently, wherespecialized skills are scarce, decentralization

172 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

reduces the availability of expertise at the locallevel. Rarely does it provide an adequate fiscalbasis for the transferred responsibilities. This cre-ates new challenges in municipal management.It requires local authorities to acquire technicalcompetencies to fulfill new responsibilities. Itgives municipalities the incentive to incorporateurban transport in their city development strate-gies and to consider the tradeoffs and synergiesamong sector policies. It may also encourageautonomous municipalities within a city regionto collaborate in the development of metropol-itan strategies and may provide a new founda-tion for more-integrated financing arrangements.The main policy prescriptions to exploit theseopportunities include the following:

• Metropolitan regions should develop met-ropolitan-scale transport planning andfinancing arrangements.

• Cities should prepare human resource devel-opment plans for their new responsibilities.

• Cities should develop integrated, fungible,urban transport funding arrangements.

• Cities should develop integrated road useand public transport charging strategies tomake the urban transport sector financiallysustainable.

PROGRESS SO FARThe Bank has already begun to respond to thesechallenges.

• Institutional development. Functional coor-dination and human resource developmenthave been successfully pursued in manyurban transport projects. For example,regional transport coordination institutionshave been reestablished in several of themajor Brazilian conurbations in associationwith Bank-funded mass transit investments;similar efforts are being undertaken to re-establish an effective metropolitan planningfunction in Metro Manila and in Dakar.

• Human resource development. The Bankhas attempted to provide technical assis-tance and training for new responsibilities.

For example, study tours to acquaint munic-ipalities with franchising experience abroadhave become a common feature of publictransport projects in Eastern Europe andCentral Asia.

• Financing arrangements. Modally integratedticketing systems are being supported inBrazil. Where there are multiple municipal-ities in a metropolitan region, the creation ofa single-purpose authority or the poolingof metropolitan transport funding may makeresources more fungible between modes,as well as give a new incentive to the devel-opment of new road taxation or chargingsystems to meet the metropolitan-levelneeds. In some of the transition economiesof the former Soviet Union, efforts are beingmade to ensure that adequate allocationsare made to urban roads from national roadfunds.

THE AGENDAIntegrated planning of urban transport infra-structure and services is important in any setting.As responsibility for the management of urbanservices is decentralized to the cities, such inte-gration becomes more feasible. For integrationto be successful, there must be a conscious effortto develop a strategic structure plan to controlland use and to coordinate the development ofurban services within the long-term developmentstrategy. The Bank can contribute to improvingthe effectiveness of decentralization in the fol-lowing ways:

• Support the formulation of city develop-ment strategies and the technical instru-ments, such as land use and developmentcontrols, that implement the plans.

• Assist institutional reform as far as possibleby technical assistance components in tra-ditional Project and Sector Adjustment loaninstruments, as well as by use of newer instru-ments (such as the Learning and InnovationLoan [LIL] and programmatic lending instru-ments) to give a broader and more continu-ous support to reform programs.

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 173

• Continue to encourage the developmentof sustainable metropolitan and municipaltransport financing, both by encouraginginstitutional reform within urban transportprojects and, more generally, by program-matic lending to municipalities.

• Where governance is not good, or wherejurisdictional or functional fragmentation isparticularly damaging, support metropoli-tan-level single-purpose government organ-izations with appropriate integrated urbantransport funding arrangements.

MOBILIZING PRIVATEPARTICIPATION

Inability to support the increasing fiscal burdenof public sector operations, associated with adesire to secure investment off-budget, has ledmany countries, including those in traditionalpublic sector–dominated areas such as theformer Soviet Union, to turn to the privatesector for financing and operation of transportinfrastructure and services. But they have oftendone so without realizing, or adequatelyaddressing, the very important role that theystill must play. Fares and tolls have often beenset without regard to the sustainability of theservice to be provided. Where there is an effec-tive competitive alternative, the greatest degreeof freedom in fare setting is desirable. However,the grant of concessions or franchises usuallyconfers a degree of monopoly power, whichmight in some cases result in profits being max-imized at relatively high prices and low utiliza-tion. The exclusion of poorer users may reduceoverall welfare even though it increases prof-itability. Ad hoc opportunism in the sanction-ing and support of private sector–financedinfrastructure in developing-country cities hasoften led to poor physical and commercial coor-dination of developments (as in Bangkok andKuala Lumpur) and significant, unplanned, con-tingent costs falling on the public authorities(as in Manila). The main policy prescriptions toaddress these deficiencies are the following:

• Government should prepare framework lawsand model contracts for infrastructure con-cessions and competitively tendered serv-ice franchising.

• Cities should prepare strategic infrastruc-ture plans with which private sector devel-opments should be consistent.

• Cities should establish public transport plan-ning and procurement units, independentof operating agencies.

• Cities should develop criteria for appraisingthe public financial contribution to public-private participatory schemes, such as con-cessions.

PROGRESS SO FAR The Bank has done much to encourage privateparticipation in the sector, attempting to assistin the development of efficient and equitableforms for private participation. The main areasof development have been as follows.

• Infrastructure concessioning. The Bank hasbeen instrumental in assisting the conces-sioning of some urban railways to the privatesector, particularly in Latin America. That expe-rience has shown that urban mass transit sys-tems can benefit from private managementand that front-end-loaded contributions to aprivate-public participatory arrangement canimprove service quality and reduce thebudget burden. Urban motorway concessionshave also increased capacity, but have cre-ated problems where they are not establishedwithin a broad city infrastructure developmentplan.

• Urban road–passenger transport franchis-ing. The Bank has for many years advocatedreliance on competitive private sector supply,often in the form of an increasing role for theinformal sector and small vehicles. However,different experiences in countries such asChile, Peru, and Sri Lanka have shown that,particularly in large conurbations where envi-ronmental and congestion externalities aremost significant, well-regulated competition“for the market” may better meet the aspi-

174 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

rations of the city than free competition “inthe market.” The Bank has extensively usedproject preparation funds to acquaint deci-sionmakers in the transitional economieswith a range of competitive market forms,and with the administrative experience ofcountries that have successfully employedmarket forms.

• Bus finance. Past lending to public sectorbus companies has assisted in meetingimmediate supply problems (as in Russia),but has often failed to ensure the sustain-ability of that improvement. Consequently,recent projects in Kazakhstan and Uzbekistanhave concentrated on finding an institutionalbasis on which affordable public transportsupply can be maintained, particularlythrough the introduction of competitive pri-vate sector supply. Direct financing of publicsector vehicles tends to undermine that com-petitive initiative, as well as to confront seri-ous problems of vehicle specification forprocurement. For that reason the Bank hastried to encourage the development of sup-plier finance or commercial leasing arrange-ments, which can be accessed by all formsof supply agency, but which require them tobe able to meet the full costs of vehiclefinance and maintenance.

• Affordable fares. The Bank has supportedvarious devices to ensure that private par-ticipation does not make public transportunaffordable, including financing initial gov-ernment contributions to concession agree-ments and public specification of fare tablesin contract documents in Argentina, usingfare level as a selection criterion in tendersin Uzbekistan, or subjecting tariff or profitlevels to an independent regulator, as in sev-eral rail concessions.

• Strengthening the planning context. TheBank has addressed this, ex ante, throughthe development of coordination arrange-ments in the context of urban rail develop-ments in Brazil and, ex post, through thefinancing of modal interchange betweenbuses and light rapid transit in Manila.

• Public-private participation. Privatelyfinanced urban transport infrastructure canyield benefits not only to its users—whoshould normally be willing to pay for the ben-efits accruing to them—but also to nonusers,whose travel conditions are relieved by thetransfer of some traffic to the new facilities.This highlights the importance of treatingthese developments as genuine public-private partnerships to which publ icsector–funding contributions are entirelyappropriate. This contribution might be newcapital, existing complementary infrastruc-ture, or even operating support. It also pointsto the possibility of combining Bank invest-ment in the public sector element with anInternational Finance Corporation (IFC) par-ticipation in the private finance. This requiresattention to both the institutional arrange-ments for and the evaluation of private sectorparticipation in infrastructure finance. TheBank has funded public sector contributionsto concessioning arrangements in Argentinaand Brazil.

THE AGENDAWhile World Bank Institute (WBI) training and dis-semination efforts in infrastructure finance haveraised the awareness of many problems, muchremains to be done to consolidate the develop-ments of private participation in urban transport.Concession contract design deficiencies have cre-ated some severe problems (for example, theBangkok Second Expressway). The need to rene-gotiate obsolete contracts has been a source ofmuch friction in Argentina. Regulatory provisionshave often been inadequate for both infrastructureconcessions and service franchises. It is also impor-tant that private sector developments should beconsistent with a city strategy for physical devel-opment, and justified in system cost-benefit, aswell as private financial, terms. The Bank can con-tribute to more-effective private sector partici-pation in urban transport in the following ways:

• Disseminate good practice in franchisingand concessioning through WBI activities,

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 175

including newly developed distance-learn-ing packages.

• Further develop and tailor guaranteearrangements to the special needs arisingin transport concessions.

• Only be involved in finance of vehicleswhere it contributes to a desirable reform,either in the introduction of cleaner or safertechnologies or in sustainable competitivesupply arrangements.

• Support the development of commercialfinancing arrangements for vehicles andseek means to provide microcredit to smalloperators and for a wide range of vehicletype.

• Continue to encourage the developmentof forms of concession and franchise con-tracts that ensure the benefits of competi-tion are passed on to poor users and thatenvironmental standards are maintained.

• Finance the development of skills in pro-curement and regulation, through both proj-ect and reform program loan financing.

• Continue the effort to improve the formsof competition in service provision, withparticular emphasis on the LIL instrumentas a vehicle to support experimentation andpreparation, but also through the incorpo-ration of institutional audits and analysis asa routine part of the preparation of all lend-ing activities.

• Support the physical and commercial coor-dination of private and public transportfacilities both through specific investmentsto deal with defects in existing systems andthrough technical assistance to establishand to staff planning and coordinatingagencies.

• Support the effective extension of privatesector participation in urban transport infra-structure finance by requiring that suchdevelopments be consistent with an estab-lished city development plan, and thatpublic sector contributions, in whateverform, be subject to an economic appraisalsimilar to that undertaken for purely publicprojects.

INCREASING TRANSPORT SAFETYAND SECURITY

Road accidents already represent a global pan-demic, and deteriorating security is a growing prob-lem, particularly in Latin America. Absence ofsidewalks, inadequate lighting, and poor segre-gation of motorized and nonmotorized traffic makepedestrians and bicyclists vulnerable both to traf-fic accidents and to criminal attack. Dangerouspublic transport operating behavior has been asignificant contributor to road traffic accidents insome countries in the absence of effective polic-ing of traffic behavior. The poor suffer dispropor-tionately from these deficiencies. Transport safetyand security in developing countries suffer because,although they require wide collaboration amongagencies (transport, health, police, and so on), theyare often “nobody’s baby.” To overcome this it isnecessary both to develop appropriate interagencycoordination mechanisms and to develop the nec-essary skills. The main policy prescriptions toaddress these deficiencies are the following:

• Cities should identify and empower a unit ofits administration, or quasi-independentagency, with police involvement, to be respon-sible for safety and security in urban transport.

• All urban transport infrastructure and trafficmanagement plans should be subject to amandatory safety audit before approval.

• Traffic police should be involved in design oftraffic management schemes, but should besubject to instructions from the traffic man-agement authority on the objectives of traf-fic management policy and enforcement.

• All urban transport franchises or operating per-missions should carry strict requirements forsafe operating practice and be should berescindable for nonobservance of that practice.

PROGRESS SO FAR Attempts to address the problems so far includethe following:

• Infrastructure design. Recent Bank-financedprojects have provided segregated facilities

176 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

for cyclists in Lima, Accra, and Manila, as wellas tried to maintain good cycling facilities inChinese cities. That concern must now bemainstreamed into all urban projects. Insome cases (for example, in many residen-tial areas), the best solution may not be com-p le te sepa ra t ion o f moto r i zed andnonmotorized traffic but the adoption ofmeasures to “calm” motorized traffic tospeeds at which it is not a threat to pedes-trians or cyclists. Protection of pedestrianshas also been pursued through the financ-ing of sidewalks in many cities (for example,Accra, Nairobi, and Dakar).

• Institutions. Despite the reputational riskissues of World Bank association with invest-ments in policing in many countries, the Bankhas recognized that it is unlikely that traffic-safety problems can be adequately addressedwithout substantial police involvement. Mostrecently, in conjunction with an urban trans-port project in Moscow, the Bank has foundit necessary to further refine the criteria foracceptability of such involvement, and toensure that the police are adequately involvedin traffic system design and trained to per-form traffic management functions in an effi-cient manner consistent with the city trafficsafety and traffic management strategies.

• Regulation, market structure, and safety. Insome cases, such as in Lima, deterioration inboth safety and security has been associatedwith market liberalization. In others, such asthe concessioning of the Buenos Aires sub-urban railways, the passenger-security situation has been improved by privateinvolvement. More understanding is requiredof the relationship between safety, security,and market structure in urban passengertransport, but there already appears to be aclose association between the quality of eco-nomic regulation and the safety and securitysituation.

THE AGENDA Personal security is only partly a transport sectorphenomenon per se, but it has particularly seri-

ous impacts on the sector. In both areas, the chal-lenge to the Bank is to help clients develop waysof confronting these pressures in the context ofan increasingly fragmented public transportsupply situation. The Bank can contribute to thedevelopment of a concerted effort to improvetransport safety and security in the following ways:

• Require adequate safe provision for pedes-trians, including mobility-impaired pedes-trians, and NMT (especially bicycles) in allnew road investment, rehabilitation, andtraff ic management intervent ions .Moreover, safety should not be used as ajustification for greatly inconveniencingthese categories of road user to facilitateeasier flow of motorized vehicles.

• Support broader institutional involvementin transport-related security and traffic-safety issues, including improved coordi-nation between traffic management andpolice agencies.

• Further refine the criteria under which spe-cific items of investment in policing areappraised, paying special attention to theneed to invest only in facilities and equip-ment that cannot easily be converted topurposes of civil repression.

• Attempt to ensure that the safety of opera-tion of vehicles and the security of passen-gers, including the mobility impaired, areproperly addressed in public transport reformmeasures, particularly where existing monop-olies are replaced by multiple operators.

• Avoid, or adopt appropriate mitigatingmeasures against, any serious potential risksto the environment and public health andsafety associated with projects or programs.

PROTECTING THE ENVIRONMENT

Although environmental impacts of transport arereaching the top of the agenda in many mega-cities, municipalities still tend to ignore the envi-ronmental implications of their strategies untiltoo late in the process of growth, to ignore global

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 177

warming issues at any level, and to fail to recog-nize the distributional implications of environ-mental impacts. One of the lessons of experienceis that while traffic management can improve net-work performance, it cannot be a substitute foradequate capacity provision and network plan-ning in rapidly growing cities. The weakest linkin the development of the strategic position hasconcerned the efforts to advocate demand man-agement, particularly through road pricing orother surrogate devices to ensure that users paythe full social costs of road use. The general strat-egy that is recommended for protecting the envi-ronment is a combination of better managementof urban transport operations in the short term,with technological improvement and urban struc-tural change in the longer term. The main policyprescriptions to achieve this are the following:

• National programs should be established,where necessary, to eliminate leaded gaso-line and to improve or replace the technol-ogy of gross emitters of particulate matterwhere that is cost-effective.

• Cities should establish a traffic managementdepartment or agency with responsibility formeasures to reduce environmental impactof traffic, as well as measures to improve itsefficiency.

• Municipal action should improve publictransport as a high priority.

• Traffic system management and trafficdemand management programs should bedeveloped.

PROGRESS SO FAR The ways in which these prescriptions have beenpursued in Bank operations include the following:

• Supporting technological development.Because much of the necessary technologicaldevelopment occurs in the private sector, itmay fall outside the ambit of traditional Banklending. A number of new instruments mayhelp to overcome this limitation. The IFC mayinvest or lend for innovations of commercialpotential. In addition, new instruments are

becoming available in some areas. For exam-ple, technological development of means ofimproving fuel efficiency in transport mayreceive funding, both through OperationalProgram 11 of the Global Environment Facility(GEF) and through investments undertakenby the Prototype Carbon Fund.

• Traffic management. The Bank has put muchemphasis on traffic management in its pasturban transport strategies. It has achievedconsiderable success in improving traffic flowin countries such as Brazil, Chile, and Korea,and has included similar measures in currentprojects in Beirut (Lebanon) and Moscow,including substantial investments in trafficcontrol equipment. Among the most recentdevelopments has been an increasingemphasis on providing for NMT and pedes-trians, as in Lima and Manila. The Bank isalso administering GEF grants in process orin preparation, aiming at improved air qual-ity through traffic management measures inLima, Manila, and Santiago.

• Road infrastructure. Road-related expendi-tures account for about one-third of theBank’s urban transport portfolio. About one-half of that has been on capacity rehabilita-tion rather than expansion, with even thateffort concentrated on the roads that carrythe main public transport flows, as in theKyrgyz Republic urban transport project. Forlarger established cities, it is recognized thatit is neither possible nor desirable for coun-tries to attempt to invest their way out ofroad congestion. That said, it is clear thatthe Bank has not withdrawn entirely from thesupport of urban road infrastructure invest-ment , no r does i t i n tend to do so .Nevertheless, it is critical that any Bank lend-ing for urban road investment be conditionalon the implementation of effective policiesof traffic management and restraint.

• Demand management, prices, and taxes.Despite disappointing past experience, thedevolution of responsibilities to the cities inthe context of severe fiscal pressure,together with a developing interest and

178 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

experience with road pricing in some indus-trialized countries, makes it now opportuneto make the development of road-pricingmechanisms a priority.

• Bus mass transit. The concept of physicallyseparating buses from worsening traffic con-gestion was supported by the Bank in proj-ects in Brazil, Chile, and elsewhere. Therecent revival of bus segregation as a moreextensive mass transit system, first intro-duced in Curitiba, is now being supportedin Bogotá and is being considered in Lima.

• Rail mass transit. Previous policy statementsand actions taken have given the impressionthat the Bank was resolutely opposed to anyfixed-rail investment in urban areas for fearthat investment in such large, expensive proj-ects would preempt municipal funds, whichmight yield higher social returns elsewhere.The portfolio review shows how substantiallythat position has been amended. The posi-tion that is now adopted is that where urbanrail systems already exist, it is usually desir-able to sustain them (for example, the proj-ects in Argentina and Brazil) and that newurban rail investments must be seen in thewider context of a long-term urban devel-opment strategy. What remains of the tra-ditional position is a determination that railinvestments must be fiscally sustainable.What has been added is a determinationthat such developments must be integratedinto a comprehensive urban transport strat-egy and that arrangements should includephysical and fare integration between modesto ensure that the poor are not excludedfrom or disadvantaged by the investments.At the technical level, this is associated withrecommendations both about the way inwhich the distributional consequences ofurban rail investments should be estimatedand about the way in which the economicbenefits should be evaluated.

THE AGENDA The Bank retains its strong commitment to envi-ronmental improvement. It is recognized that this

cannot be approached as a technically and insti-tutionally separate activity but must be integratedin an urban transport policy with other objectiveswith which it has to be reconciled. The Bank canassist countries and city governments to maketheir urban transport sectors more environmen-tally sustainable in the following ways:

• Emphasize the importance of road planningfor the movement of people and not just ofvehicles. Increase the emphasis on the needsof public transport passengers, pedestri-ans, and users of nonmotorized modes oftransport in urban road operations.

• Advocate a balanced strategy in urbanareas, combining the provision of a neces-sary basic amount of road capacity with aprogram of traffic management, demandrestraint, and public transport development;link any urban road investments to the intro-duction of supporting traffic managementand pricing measures to ensure thatplanned environmental and efficiency ben-efits are widely distributed.

• Attempt to ensure that the methodologiesfor financial and economic appraisal ofinvestments are consistent and compara-ble between modes.

• Renew efforts to secure the introduction ofeffective measures of demand restraint,through both price and nonprice instruments.Where compliance with environmental stan-dards requires other management measuresto be associated with investments, requireevidence that the management measuresare being implemented before the invest-ment funding is committed.

• Support the introduction or extension ofbus rapid transit systems, linking it to areform of public transport supply whereappropriate, especially where this can beshown to enhance the environment.

• Support the exploitation of the potentialof urban rail, particularly where the basicinfrastructure already exists, but continueto be cautious in ensuring that the appraisalof any such investments is realistic and that

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 179

the long-term fiscal burden of any invest-ment in urban rail is properly assessed andseen not to deflect from other, more ben-eficial investments.

• Search for and support technological devel-opments to increase efficiency of fuel uti-lization in transport, but recognize thattechnological development in transport isprimarily a private sector prerogative andresponsibility, and act as agents of the GEFand the Prototype Carbon Fund.

• Ensure that reasonable allowance is madefor the effects of induced traffic in assess-ing the economic and environmentalimpacts of road infrastructure investments.

CONCLUSIONS

We turn finally to the question of how the strat-egy will be implemented, with particular empha-sis on the instruments, processes, and proce-dures to be adopted.

INSTRUMENTSThe context for this review is a general shift inWorld Bank policy on its approach to infrastruc-ture sectors. Three shifts in Bank corporate policydeserve mention:

First, there is the shift from infrastructure proj-ect finance toward knowledge building, advi-sory services, and capacity building. The LIL isa small, quick-disbursing instrument devised toenable reforms to be introduced on an experi-mental basis, and to decouple support of suchexperiments from investment lending. It isadministratively expensive, however, and hasgenerated only limited enthusiasm on the part ofboth staff and clients. In any event, institutionalreform is usually a slow, learning process, andmay need a more protracted and systematicpresence than the LIL allows. Associated withthe emphasis on knowledge building is a newemphasis on output-based aid, where supportis linked not to investment in public sector supply,but to the development of efficient procedures

of procurement of privately provided services tothe poor.

Second, there is a shift from stand-alone projectfinance toward more-programmatic lending. Thismay take several different forms, including thefollowing:

• Funding a time slice of a single-sector pro-gram based on agreement about the asso-ciated policy objectives and reforms, suchas traditional sectoral investment and adjust-ment lending

• Multisector lending aiming at the integra-tion of sector inputs within a comprehensivestrategy, such as the transport componentof the Rio de Janeiro State Private SectorDevelopment Project

• A tranched approach linked to and triggeredby the sequential achievement of reform tar-gets, as in Adjustable Program Lending, suchas the Urban Mobility Project for Senegal.

Several critical issues in this review—such asmunicipal institutional reform, regulation of con-cession contracts, local air pollution, and so on—are transport applications of more-genericproblems. An increasing proportion of the Bankassistance to reforming urban transport institu-tions may thus appear in wider projects dealingwith municipal reform or air quality over a rangeof sectors. In that context it is important to ensurethat sectoral matters continue to be dealt within a technically competent manner that does notweaken the strong professional links that havealways been so important in transferring knowl-edge and securing developing-country commit-ment at the appropriate technical and politicallevels. The rigorous World Bank loan procure-ment rules, the objective of which is to ensurefair competition and to avoid possibilities of cor-ruption, will remain in place.

Third, there is a desire to shift from conventionalto high-impact lending. This includes novel activ-ities, particularly in public-private participation,which are intended to have a high demonstra-

180 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

tion effect. For example, the Bank partial riskguarantee can protect franchisees against defaultby government in performance of their contrac-tual obligations, although in practice there havebeen no such guarantees to date. In some otherimportant aspects of urban transport policy—particularly the development of NMT and theinformal public transport sector—the critical needis the provision of relatively small amounts offinance to multiple, usually private, borrowers.The Bank’s experience with lines of credit—par-ticularly with those tied specifically to selectedsectors—has not been good. It is therefore impor-tant that the best experience of the Bank inaddressing the financing needs of small andmedium-size enterprises be incorporated in urbantransport sector developments. It is also nowbeing explicitly recognized that in some areas,the Bank has global advocacy goals that may bein advance of current client perception and pref-erences. In the urban transport sector, the Bankhas already entered such a campaign forimproved transport safety and security.

LENDING TO CITIESAs responsibility for urban transport is decen-tralized, cities are increasingly forced to maketheir own decisions about financing, includingborrowing. City credit risks can be particularlyserious because the Bank’s loans are typicallydenominated in foreign currencies and cities’ rev-enues are in local currency. Unlike some of theother multinationals,2 the International Bank forReconstruction and Development (IBRD) is con-strained by its charter from lending directly tocities, unless there is a sovereign guarantee. Incontrast, the IFC can lend only to the privatesector, and must do so without sovereign guar-antee. Cities therefore fall between the IBRD andIFC conditions as imposed on government andprivate borrowers.

The effects of this situation are complex. On theone hand, cities may face the moral hazard of anincentive to unduly risky borrowing if guaranteedby the central government. This perverse incen-tive can be countered by appropriate pricing for

the guarantee (for example, both the Polish andColombian governments have charged a 2 per-cent fee up front for the guarantee). On the otherhand, as an incentive to reform, governmentsmay offer guarantees on a competitive basis tocities undertaking reforms that will increase theirfiscal sustainability.3 Properly priced or condi-tioned sovereign guarantees may thus contributeto the achievement of the sustainability objec-tive. The most serious concern, given the Bank’semphasis on supporting effective decentraliza-tion, is that a central government may use itsguarantee power as leverage over city govern-ments of a different political persuasion from thecentral government, or may shy away from theunpleasant task of discriminating between cred-itworthy and noncreditworthy cities.

In some cases channels of finance other than director guaranteed lending to municipalities may befound to meet the needs of urban transport. TheIBRD has lent to public banks in Colombia, whichhave then on-lent to cities. In Uzbekistan a loan toa public sector leasing agency, operating within a“sunset clause,” has been used to enable vehi-cles to be made available to public and state jointstock companies in the secondary cities. The IFCcan lend to private, or mixed public-private, banks,which can on-lend on commercial terms to trans-port service suppliers. The experience of devel-oping microcredit for urban transport vehicles isnot very promising, however, and supplier credit forleased vehicles has been difficult to stimulate insome of the transitional economies where it is mostrequired; the need for a wider range of lendinginstruments remains.

The crucial issue from the Bank’s point of view iswhether the instruments available are able to for-ward the objectives of strengthening the capac-ity of the cities to become fiscally self-reliant andto access private capital markets. Some traditionallending—for example, to state-owned enterprisesfor bus replacement—may have actually discour-aged the development of that sustainability. Theemphasis is therefore shifting to the developmentof supplier credits, microcredit, and other more

MEETING THE DEVELOPMENT CHALLENGES: HOW CAN THE BANK CONTRIBUTE? 181

commercially based financing arrangements, andaway from the financing of public enterprises. TheIFC may have a role to play in assisting the estab-lishment of commercial agencies able to lend tomunicipalities, such as Infrastructure FinanceCorporation Limited in South Africa; the IFC isunlikely, however, to be involved in financing directlines of credit for municipalities.

PROCEDURES: OWNERSHIP, PARTICIPATION, AND COLLABORATIONThe Bank’s lending programs emerge from a “top-down” process, at the apex of which is the CountryAssistance Strategy or, for heavily indebted poorcountries, the Poverty Reduction Strategy Papers.These strategic documents are increasingly a jointproduct of the national government and the Bank,often supported by wider stakeholder consulta-tion. The inclusion of urban transport componentsin country programs thus depends on urban trans-port’s having been identified as a priority problemarea in national development strategy. Althoughurban transport pervades the whole urban econ-omy, its influence tends to be diffused. It is there-fore important that the urban transport–povertynexus is better understood, and that urban trans-port developments are increasingly linked both togeneral urban development activities and to otherspecific social sector developments, including theacknowledgment of women and the mobilityimpaired as important stakeholders.

Local ownership is always essential to project suc-cess. In the rural context, civil society is increas-ing ly be ing engaged in the p rocess o fcommunity-driven development. That is more dif-ficult in urban areas, where populations are oftenmore transitory and local institutions corre-spondingly weaker. But there are some very impor-tant policy questions on which transport-userparticipation is being increasingly fostered in theinterests of more-sensitive project or programdesign. These include the following:

• Design and implementation of resettlementprovisions. In the Mumbai Urban Transport

Project, local slumdwellers’ associations areengaged in defining strategy and selectinglocations and phasing of resettlement pro-visions.

• Design and implementation of paving pro-grams in poor areas. Citizens of the munic-ipality of Villa el Salvador, on the outskirtsof Lima, have developed their own localroad-paving strategy and scheme. In thepreparation of an urban transport projectfor Lima, inhabitants of poor areas (“pueb-los jovenes”) participate in the selection ofroad links for paving to give vehicular accessto poor areas.

• Service quality and fare tradeoff. Social sur-veys of public transport passengers are anessential element of public transport proj-ect design. Such surveys in Cali, Colombia,showed that the proposed expenditure on ametro system was not perceived as helpingthe poor, despite the politicians’ claims.Conversely, in Uzbekistan and the KyrgyzRepublic, user surveys showed a muchgreater concern with the availability of serv-ice than politicians had asserted.

• Safety and security policy formulation. Fartoo little is usually done to establish howand where transport users are most vul-nerable, both to road accidents and, moresignificant, to personal violence on trans-port. In the project under preparation inLima, a detailed survey of women hasdemonstrated the significant impact of on-vehicle violence and harassment on theactivity patterns of women and hence theirultimate well-being.

In undertaking the present review, the Bank hasattempted to involve nongovernmental organi-zations and individual members of civil society,as well as governmental and professional bodies,to respond to the proposals being made.

The Comprehensive Development Frameworkalso raises the question of the comparativeadvantage of the Bank in each specific area ofactivity. The Bank is already frequently involved

182 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

in collaboration with other international lend-ing and donor organizations, and some of thoseinstitutions have financed background studiesfor this review. The European Union has recentlyexpressed a parallel wish to improve collabora-tion on transport development matters. Typically,collaboration involves joint or parallel financingof projects. In some cases, however, such as theurban Sub-Saharan Africa Transport Program,the collaboration extends to the program level.One of the main purposes of publishing a strat-egy review such as this is to solicit commentfrom peer institutions with the aim of increas-ingly aligning the positions of the institutionson major development strategy issues. The eco-nomic, environmental, and social problems ofurban transport are large, and in many countriesthese problems continue to grow. The issuesneed, and deserve, the concentrated and coher-ent attention of the whole development com-munity.

NOTES

1. The concerns expressed in the Voices of thePoor studies were approximately equally dividedbetween health, education, and infrastructureissues (Narayan 2000b).

2. For example, the European Bank forReconstruction and Development is able to lendto cities without sovereign guarantee if the con-ditions merit it, while the European InvestmentBank, although requiring a guarantee, can accepta guarantee purchased by the borrower from acommercial bank.

3. For example, the government of the RussianFederation offered a guarantee for funds used fromthe First Urban Transport loan to secondary cities,conditional on the cities promising to achieve apreset level of cost recovery deemed necessary tomake their operations sustainable. In the event,though, the city finances were not in good condi-tion and yet were able to meet the terms.

APPENDIX: URBAN TRANSPORT PORTFOLIO IN THE WORLD BANK 183

APPENDIX: URBAN TRANSPORTPORTFOLIO IN THE WORLD BANK

Fiscal YearApproved

2000a

1995

2002

2001

2000a

1996

1999

1997

2001a

2002

2002

2000a

1996

19941998

1998

1999

2001

Country

Africa Region

Benin

Burkina Faso

Cameroon

Chad

GhanaKenya

Malawi

Mali

Mauritania

Mauritius

Nigeria

Senegal

Tanzania

TogoZambia

East Asia andPacific Region

China

China

China

Status

Ongoing

Ongoing

Under pre-parationOngoing

OngoingOngoing

Ongoing

Ongoing

Ongoing

Under pre-paration

Under pre-paration

Ongoing

Ongoing

OngoingOngoing

Totals

Ongoing

Ongoing

Ongoing

Name of Project

First DecentralizedCity Management

(phase 1)Urban Environment

ProjectDouala Infrastructure

ProjectTransport Sector

Strategy (PAProNaT)

Urban 5 (phase 1)Urban Transport

Road Maintenance &Rehabilitation

(ROMARP)Urban Development& Decentralization

Urban DevelopmentProgram

Transport Sector

Lagos UrbanTransport

Urban MobilityImprovementUrban SectorRehabilitation

Urban DevelopmentNational Road

Guangzhou CityCenter Transport

Liaoning UrbanTransport

Urumchi UrbanTransport

UrbanTransport

21.8

9.1

86.0

2.2

5.0155.0

4.0

30.0

17.0

70.0

140.0

132.0

32.0

26.240.0

770.3

586.0

384.0

250.0

Numberof Citiesin Project

3

2

1

1

2326

2

10

13

4

1

1

8

13

1

3

1

ProjectTeam

Leader

Larbi

Ouayoro

Inal

Schliess-ler

BahalMorrell

Brushett

Inal

Inal

Ghzala

Schelling

Bultynck

Carroll

CaveBrushett

Scurfield

Scurfield

Dotson

Project Cost (US$M)

UrbanTransport

19.2

6.8

38.0

1.6

2.0115.0

4.0

25.4

14.0

3.0

100.0

98.0

30.0

11.011.0

479.0

200.0

150.0

100.0

Total

25.5

38.0

38.0

67.0

11.0115.0

30.0

80.0

70.0

20.0

100.0

98.0

105.0

13.070.0

880.5

200.0

150.0

100.0

WB Loan (US$M)

Principal UrbanTransport Thrust

Urban road rehabilitation &drainage

Urban road improvement &rehabilitation

Urban road rehabilitation &management

Road safety improvement

Urban road improvementTransport infrastructure and

managementUrban road rehabilitation &

maintenance

Urban road rehabilitation &maintenance

Transport infrastructureimprovement & road

constructionEnvironmental infrastructure

improvementInstitutional reform, road

rehabilitation & bus operationimprovement

Infrastructure improvement andprivate sector supportRoad rehabilitation &

constructionUrban road & drainage

Urban road rehabilitation &support of urban road

management

Road investment & maint., trafficmgmt./safety, environment,

public transportRoad investment & maint., traffic

mgmt./safety, environment,public transport

Road investment & maint., trafficmgmt./safety, environment,

public transport

Total

28.3

49.7

86.0

91.1

22.0155.0

39.0

93.5

100.0

100.0

140.0

132.0

141.0

29.0454.0

1,660.6

586.0

384.0

250.0

A.1 URBAN TRANSPORT PROJECTS UNDER PREPARATION AND IMPLEMENTATION AS OF FEBRUARY 26, 2002

184 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

WB Loan (US$M)

A.1 URBAN TRANSPORT PROJECTS UNDER PREPARATION AND IMPLEMENTATION AS OF FEBRUARY 26, 2002 (CONTINUED)

Project Cost (US$M)

Country Name of ProjectFiscal YearApproved Status Total Total

UrbanTransport

UrbanTransport

Principal UrbanTransport Thrust

Numberof Citiesin Project

ProjectTeamLeader

Shijiazhuang UrbanTransport

Wuhan UrbanTransport

Strategic UrbanRoads

Infrastructure

Strategic Urban Road Infrastructure II

(Java)

Pusan UrbanTransport

ManagementMetro Manila Urban

Transport Integration

Bangkok Air QualityManagement

Urban TransportImprovement

Armenia TransportProject

Urban Transport

Municipal ServicesUrban Transport

Bridge Rehabilitation

Moscow UrbanTransport

Urban Transport

Second MunicipalDevelopment

China

China

Indonesia

Indonesia

Republic ofKorea

Philippines

Thailand

Vietnam

Europe andCentral AsiaRegion

Armenia

Kyrgyz Rep.

LatviaRussian

FederationRussian

FederationRussian

FederationUzbekistan

Latin Americaand theCaribbeanRegion

Argentina

2001

2003

1996

2003

1994

2001

2002

1998

2000

2001

19951995

1997

2001

2000

1995

Under pre-paration

Concept

Ongoing

Under pre-paration

Ongoing

Under pre-paration

Under pre-parationOngoing

Totals

Ongoing

Ongoing

OngoingOngoing

Ongoing

Under pre-parationOngoing

Totals

286.0

500.0

168.0

150.0

332.0

87.0

80.0

47.0

2,870.0

47.0

24.2

45.4391.0

466.0

100.0

31.4

1,105.0

600.0

286.0

500.0

168.0

150.0

332.0

87.0

80.0

47.0

2,870.0

0.5

24.2

15.0391.0

93.0

100.0

31.4

655.1

186.0

100.0

200.0

87.0

150.0

100.0

56.0

80.0

43.0

1,266.0

40.0

22.0

27.3330.0

350.0

60.0

29.0

858.3

210.0

100.0

200.0

87.0

150.0

100.0

56.0

80.0

43.0

1,266.0

0.5

22.0

15.0330.0

69.8

60.0

29.0

526.3

65.0

Road investment & maint.,traffic mgmt./safety,

environment, public transportRoad investment & maint.,

traffic mgmt./safety,environment, public transportReducing traffic bottleneck,

decentralization of responsibility of urban

transportReducing traffic bottleneck,

decentralization of responsibility of urban

transportEffectiveness of an urban rail

transit & policy reforms

Improving integrated access topublic Iransport & NMT

Reduce local emissions

Operational efficiency & safetyimprovement

Urban transport study

Urban road rehabilitation &finance

Bus service improvement in RigaExpand fleet & reform public

transport operationsBridge rehabilitation in Moscow

Traffic management + associatedinstitutional developmentPublic transport reform &

leasing of buses

Road paving

1

1

8

TBA

1

12 citiesand 5

mun. inmetroManila

1

1

1

3

114

1

1

5

Many

Scurfield

Dotson

Scouller

Dotson

Burning-ham

Burning-ham

Capan-nelli

Burning-ham

Talvitie

Podolske

BalkindPodolske

Podolske

Podolske

Crochet

MiguelMercado-

Diaz

APPENDIX: URBAN TRANSPORT PORTFOLIO IN THE WORLD BANK 185

WB Loan (US$M)

A.1 URBAN TRANSPORT PROJECTS UNDER PREPARATION AND IMPLEMENTATION AS OF FEBRUARY 26, 2002 (CONTINUED)

Project Cost (US$M)

Country Name of ProjectFiscal YearApproved Status Total Total

UrbanTransport

UrbanTransport

Principal UrbanTransport Thrust

Numberof Citiesin Project

ProjectTeamLeader

Buenos Aires UrbanTransportPollution

ManagementDecentralization

Belo HorizonteMetropolitan

TransportRecife TransportDecentralization

São Paulo IntegratedUrban Transport

(BF)Rio de Janeiro Mass

TransitSalvador Urban

TransportFortaleza Urban

Transport

São Paulo Line 4Recife UrbanUpgrading

Regulatory Reform TA

Bogotá UrbanServices

Medium-Size CitiesUrban Transport

Second Air Qualityand Transport

ProjectLima Urban Transport

Urban Transport

Caracas SlumUpgrading

Transport TechnicalAssistance

Urban TransportDevelopmentFirst MunicipalInfrastructure

Argentina

Argentina

Bolivia

Brazil

Brazil

Brazil

Brazil

Brazil

Brazil

BrazilBrazil

Colombia

Colombia

Mexico

Mexico

Peru

Venezuela,Rep. Bol. de

Venezuela,Rep. Bol. de

Middle Eastand NorthAfrica Region

Algeria

Lebanon

Lebanon

1997

1998

2001

1995

1995

1998

1998

1999

2001

20022002

1997

2003

1993

2004

2003

1993

1998

2002

2002

2000

Ongoing

Ongoing

Under pre-parationOngoing

Ongoing

Ongoing

Ongoing

Ongoing

Ongoing

OngoingUnder pre-parationOngoing

Under pre-paration

Ongoing

Under pre-paration

Under pre-paration

Ongoing

Ongoing

Totals

Ongoing

Under pre-parationOngoing/

Supervision

364.0

36.0

67.0

197.0

204.0

95.1

316.0

308.0

176.0

1,800.0100.0

33.0

332.0

471.0

100.0

94.0

220.0

153.0

5,066.1

11.3

115.0

100.0

364.0

15.0

3.0

197.0

204.0

95.1

316.0

308.0

176.0

1,800.022.0

2.0

136.8

471.0

TBA

94.0

220.0

30.0

4,639.9

3.9

115.0

100.0

200.0

18.0

54.0

99.0

102.0

45.0

186.0

150.0

85.0

209.056.0

13.0

150.0

200.0

70.0

35.0

100.0

60.0

1,832.0

8.7

65.0

79.0

200.0

7.5

2.4

99.0

102.0

45.0

186.0

150.0

85.0

209.012.3

0.8

68.0

200.0

TBA

35.0

100.0

11.6

1,513.6

2.8

65.0

79.0

Support rail concessions &institutional reform

Air pollution

Road rehabilitation & paving

Rehabilitate & transfercommuter rail from federal to

local governmentRehabilitate & transfer

commuter rail from federal tolocal government

Build link to integrate commuterrail systems

Rehabilitate & concession trains;improve integration

Build new metro line (BOT)

Rehabilitate & transfercommuter rail from federal to

local governmentBuild new metro line (BOT)

Road & walkway widening &paving

TA for busways, metro, &suburban rail

Feeder routes, traffic mgmt.,street maint. & institutional

reformUrban transport systems in

medium-size citiesAir quality mgmt. & urban

transport mgmt.

Busway & air pollution;improved accessibility to low

income areas & safetyStreet and traffic improvement,

institutional strengtheningRoad & walkway widening &

paving

Institutional reforms, assistanceto metro concessioning

Roads, traffic and parkingmanagement, transport planningMunicipal roads, retaining walls,

sidewalks, municipal capacity

1

1

20+

1

1

1

1

1

1

11

3

1

12

TBA

1

10

1

1

1

Menckhoff

Tlayie

Rojas

Rebelo

Rebelo

Rebelo

Rebelo

Rebelo

Rebelo

RebeloV. Serra

Challa

Persaud

James

Vegera

Guitink

Rebelo

Cira

Loir

Feghoul

Benouahi

186 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

WB Loan (US$M)

A.1 URBAN TRANSPORT PROJECTS UNDER PREPARATION AND IMPLEMENTATION AS OF FEBRUARY 26, 2002 (CONTINUED)

Project Cost (US$M)

Country Name of ProjectFiscal YearApproved Status Total Total

UrbanTransport

UrbanTransport

Principal UrbanTransport Thrust

Numberof Citiesin Project

ProjectTeamLeader

Transport SectorInvestment (APL

Phase I)Transport SectorInvestment (APL

phase II)

Dhaka UrbanTransport

Municipal Services

Mumbai UrbanTransport

Tamil Nadu UrbanDevelopment - II

Tunisia

Tunisia

South AsiaRegion

Bangladesh

Bangladesh

India

India

1998a

2001a

1999

1999

2002

1999

Ongoing

Ongoing

Totals

Ongoing

Ongoing

Under pre-parationOngoing

Totals

117.2

57.0

400.4

234.2

154.0

838.0

205.0

1,431.2

117.2

39.3

375.4

234.2

10.0

838.0

55.0

1,137.2

82.0

37.6

272.3

177.0

139.0

485.0

105.0

906.0

82.0

26.6

255.4

177.0

9.0

485.0

28.2

699.2

Transport services reforms

Transport Services reforms

Urban transport servicesimprovement

Urban road rehabilitationprogram

Sustainable public transport(rail/bus)

Urban infrastructure servicesimproved

1

1

1

4

1

…

Loir

Loir

Qureshi

Kamkwa-lala

Swamina-tha

Suzuki

Note: The data are based on the project appraisal document at the time of the approval. Whenever the detailed information did not exist in the project appraisaldocument, the values were estimated.a. Adaptable Program Lending, where “lending” refers to an International Bank for Reconstruction and Development loan or International Development Association credit.

Source: World Bank project files.

APPENDIX: URBAN TRANSPORT PORTFOLIO IN THE WORLD BANK 187

Africa Region

Benin First Decentralized 2000a X XCity Management

Burkina Faso Urban Environement Project 1995 XCameroon Douala Infrastructure Project 2002Chad Transport Sector Strategy 2001 X X X

(PAProNaT)Ghana Urban 5 (phase 1) 2000a

Kenya Urban Transport 1996 XMalawi Road Maintenance & 1999 X

Rehabilitation (ROMARP)Mali Urban Development & 1997 X

DecentralizationMauritania Urban Development 2002a X X

ProgramMauritius Transport Sector 2002 X X X X XNigeria Lagos Urban Transport 2002 X X X X X XSenegal Urban Mobility 2001a X X X X X X X

ImprovementTanzania Urban Sector Rehabilitation 1996 X XTogo Urban Development 1994 XZambia National Road 1998 X X X

East Asia and Pacific Region

China Guangzhou City 1998 X X X X X XCenter Transport

China Liaoning Urban Transport 1999 X X X XChina Urumchi Urban Transport 2001 X X X X X XChina Shijiazhuang Urban 2001 X X X X X X

TransportChina Wuhan Urban Transport 2003 X X X X X XIndonesia Strategic Urban Roads 1996 X

InfrastructureIndonesia Strategic Urban Road 2003 X X

Infrastructure II (Java)Korea, Rep. of Pusan Urban Transport 1994 X X X

ManagementPhilippines Metro Manila Urban 2001 X X X X X

TransportThailand Bangkok Air Quality 2002

ManagementVietnam Urban Transport 1998 X

Improvement

A.2 SPECIAL PROJECT FEATURES

Country Project Fisc

al y

ear

app

rove

d

Po

vert

y

HIV

Co

nces

sio

ns

and

fra

nchi

ses

Air

po

lluti

on

cont

rol

Tran

spo

rt s

afet

y

Tran

spo

rt p

lann

ing

Traf

fic

man

agem

ent

inst

itut

ions

Pub

lic t

rans

po

rt

reg

ulat

ion

Tran

spo

rt s

yste

m

inte

gra

tio

n

Imp

rove

d t

rans

po

rtfi

nanc

e

188 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

A.2 SPECIAL PROJECT FEATURES (CONTINUED)

Country Project Fisc

al y

ear

app

rove

d

Po

vert

y

HIV

Co

nces

sio

ns

and

fra

nchi

ses

Air

po

lluti

on

cont

rol

Tran

spo

rt s

afet

y

Tran

spo

rt p

lann

ing

Traf

fic

man

agem

ent

inst

itut

ions

Pub

lic t

rans

po

rt

reg

ulat

ion

Tran

spo

rt s

yste

m

inte

gra

tio

n

Imp

rove

d t

rans

po

rtfi

nanc

e

Europe and Central Asia Region

Armenia Transport Project 2000 X X X XKyrgyz Republic Urban Transport 2001 X X XLatvia Municipal Services 1995 X XRussian Urban Transport 1995 X X

FederationRussian Russia Bridge 1997 X

Federation RehabilitationRussian Moscow Urban Transport 2001 X X X

FederationUzbekistan Urban Transport 2000 X X

Latin America and the Caribbean Region

Argentina Second Municipal 1995 XDevelopment

Argentina Buenos Aires Urban 1997 X X X X X XTransport

Argentina Pollution Management 1998 XBolivia Decentralization 2001Brazil Belo Horizonte 1995 X X X X X X X X

Metropolitan TransportBrazil Recife Transport 1995 X X X X X X

DecentralizationBrazil São Paulo Integrated 1998 X X X X X X X

Urban Transport (BF)Brazil Rio de Janeiro Mass Transit 1998 X X X X X X XBrazil Salvador Urban Transport 1999 X X X X X X XBrazil Fortaleza Urban Transport 2002 X X X X X XBrazil São Paulo Line 4 2001 X X X X X XBrazil Recife Urban Upgrading 2002 XColombia Regulatory Reform TA 1997 XColombia Bogotá Urban Services 2003 X X X X XMexico Medium-Size Cities Urban 1993 X X X X X X

TransportMexico Second Air Quality and 2004 X X

Transport ProjectPeru Lima Urban Transport 2003 X X X X X XVenezuela, Urban Transport 1993 X X X X X

Rep. Bol. deVenezuela, Caracas Slum Upgrading 1999 X

Rep. Bol. de

APPENDIX: URBAN TRANSPORT PORTFOLIO IN THE WORLD BANK 189

Middle East and North Africa Region

Algeria Transport Technical 2001 X X X X XAssistance

Lebanon Urban Transport TBD X X X XDevelopment

Lebanon First Municipal 2001 X XInfrastructure

Tunisia Transport Sector 2000a X X XInvestment (APL Phase I)

Tunisia Transport Sector 1998a X X X XInvestment (APL Phase II)

South Asia Region

Bangladesh Dhaka Urban Transport 1999 X X X X XBangladesh Municipal Services 1999 XIndia Mumbai Urban Transport 2001 X X X X X XIndia Tamil Nadu Urban 1999 X X

Development II

Note: HIV = human immunodeficiency virus. For technical assistance, feature categories may overlap.a. Adaptable Program Lending, where “lending” refers to an International Bank for Reconstruction and Development loan or International Development Associationcredit.Source: World Bank project files.

A.2 SPECIAL PROJECT FEATURES (CONTINUED)

Country Project Fisc

al y

ear

app

rove

d

Po

vert

y

HIV

Co

nces

sio

ns

and

fra

nchi

ses

Air

po

lluti

on

cont

rol

Tran

spo

rt s

afet

y

Tran

spo

rt p

lann

ing

Traf

fic

man

agem

ent

inst

itut

ions

Pub

lic t

rans

po

rt

reg

ulat

ion

Tran

spo

rt s

yste

m

inte

gra

tio

n

Imp

rove

d t

rans

po

rtfi

nanc

e

190 CITIES ON THE MOVE: A WORLD BANK URBAN TRANSPORT STRATEGY REVIEW

A.3

U

RB

AN

PO

RTF

OLI

O:

PR

OJE

CT

CO

ST A

ND

WO

RLD

BA

NK

GR

OU

P L

OA

N A

ND

GR

AN

T A

MO

UN

T A

LLO

CA

TIO

N (

MIL

LIO

NS

OF

U.S

. D

OLL

AR

S)

Co

untr

yP

roje

ct

Fiscal year approved

Road and bridgerehabilitation andmaintenance

Road and bridge construction

Traffic management

Bus lanes and HOV facilities

Pedestrian and bicycle facilities

Bus, trolleybus,tramway

Metro

Total

Commuter rail

Vehicle emission control

Other investments

Other technical assistance

Technical assistanceengineering and con-struction supervision

Afr

ica

Reg

ion

Beni

nFi

rst D

ecen

traliz

ed C

ity

2000

a18

.616

.13.

23.

221

.819

.3

Man

agem

ent

Burk

ina

Faso

Urb

an E

nviro

nem

ent P

roje

ct19

959.

16.

89.

16.

8

Cam

eroo

nD

oual

a In

frast

ruct

ure

Proj

ect

2002

75.1

29.1

5.3

4.6

5.7

4.7

86.0

38.4

Cha

dTr

ansp

ort S

ecto

r Stra

tegy

20

011.

00.

70.

30.

20.

80.

60.

10.

12.

21.

6

(PA

ProN

aT)

Gha

naU

rban

5 (p

hase

1)

2000

a4.

82.

44.

82.

4

Keny

aU

rban

Tra

nspo

rt19

9625

.719

.178

.758

.46.

34.

74.

73.

539

.629

.415

5.0

115.

0

Mal

awi

Road

Mai

nten

ance

& R

ehab

ilitat

ion

1999

3.5

3.5

0.5

0.5

4.0

4.0

(RO

MA

RP)

Mal

iU

rban

Dev

elop

men

t &

1997

26.7

22.4

3.6

3.0

30.3

25.4

Dec

entra

lizat

ion

Mau

ritan

iaU

rban

Dev

elop

men

t Pro

gram

2002

a14

.310

.83.

53.

217

.814

.0

Mau

ritiu

sTr

ansp

ort S

ecto

r20

0270

.03.

070

.03.

0

Nig

eria

Lago

s U

rban

Tra

nspo

rt20

0267

.041

.19.

27.

315

.613

.34.

63.

622

.218

.921

.715

.914

0.3

100.

0

Sene

gal

Urb

an M

obilit

y Im

prov

emen

t20

00a

32.3

24.0

19.5

14.5

29.8

22.1

27.3

20.3

12.9

9.6

0.7

0.5

9.6

7.1

132.

198

.1

Tanz

ania

Urb

an S

ecto

r Reh

abilit

atio

n19

9631

.029

.11.

00.

732

.029

.8

Togo

Urb

an D

evel

opm

ent

1994

9.8

8.5

2.9

2.9

12.7

11.4

Zam

bia

Nat

iona

l Roa

d19

9833

.08.

51.

60.

54.

51.

10.

90.

940

.011

.0

East

Asia

and

Pa

cific

Reg

ion

Chi

naG

uang

zhou

City

Cen

ter T

rans

port

1998

505.

017

2.4

7.0

2.4

25.7

8.8

9.2

3.1

6.8

2.3

13.4

4.6

9.0

3.1

10.0

3.4

586.

120

0.0

Chi

naLi

aoni

ng U

rban

Tra

nspo

rt19

9929

2.0

114.

126

.010

.226

.810

.517

.46.

86.

82.

77.

22.

81.

50.

66.

52.

538

4.2

150.

1

Chi

naU

rum

chi U

rban

Tra

nspo

rt 20

0121

9.5

87.8

3.5

1.4

10.0

4.0

3.8

1.5

2.0

0.8

3.1

1.2

4.6

1.8

3.5

1.4

250.

010

0.0

Chi

naSh

ijiazh

uang

Urb

an T

rans

port

2001

219.

076

.616

.55.

89.

63.

46.

42.

215

.95.

613

.04.

52.

60.

93.

41.

228

6.4

100.

1

Chi

naW

uhan

Urb

an T

rans

port

2003

500.

020

0.0

Indo

nesia

Stra

tegi

c U

rban

Roa

ds

1996

147.

076

.116

.88.

73.

82.

016

7.6

86.8

Infra

stru

ctur

e

Indo

nesia

Stra

tegi

c U

rban

Roa

d 20

0315

0.0

150.

0

Infra

stru

ctur

e II

(Jav

a)

Kore

a, R

ep. o

fPu

san

Urb

an T

rans

port

Man

agem

ent

1994

48.2

12.2

281.

286

.22.

11.

633

1.5

100.

0

Philip

pine

sM

etro

Man

ila U

rban

Tra

nspo

rt20

0118

.912

.620

.510

.526

.018

.010

.07.

010

.07.

21.

01.

086

.456

.3

Thai

land

Bang

kok

Air

Qua

lity

Man

agem

ent

2002

80.0

80.0

Viet

nam

Urb

an T

rans

port

Impr

ovem

ent

1998

3.3

34.1

1.2

8.6

4.5

42.7

Euro

pe a

nd

Cen

tral

Asia

Re

gion

Arm

enia

Arm

enia

Tra

nspo

rt Pr

ojec

t20

000.

50.

50.

50.

5

Kyrg

yz R

epub

licU

rban

Tra

nspo

rt20

0121

.719

.62.

52.

524

.222

.1

Latv

iaM

unic

ipal

Ser

vice

s19

9515

.114

.50.

50.

515

.615

.0

Russ

ian

Fede

ratio

nU

rban

Tra

nspo

rt19

9537

8.0

318.

013

.011

.039

1.0

329.

0

Russ

ian

Fede

ratio

nBr

idge

Reh

abilit

atio

n19

9783

.762

.89.

39.

393

.072

.1

Urb

an t

rans

po

rt c

om

po

nent

Pro

ject

co

st

Wo

rld

Ban

k G

roup

loan

o

r g

rant

am

oun

t

APPENDIX: URBAN TRANSPORT PORTFOLIO IN THE WORLD BANK 191

Russ

ian

Fede

ratio

nM

osco

w U

rban

Tra

nspo

rt20

0170

.233

.422

.018

.75.

44.

897

.656

.9

Uzb

ekist

anU

rban

Tra

nspo

rt20

0027

.226

.22.

81.

91.

40.

931

.429

.0

Latin

Am

eric

a an

d th

e C

arib

bean

Re

gion

Arg

entin

aSe

cond

Mun

icip

al D

evel

opm

ent

1995

185.

065

.018

6.0

65.0

Arg

entin

aBu

enos

Aire

s U

rban

Tra

nspo

rt19

9715

7.0

55.0

90.3

71.5

3.0

1.5

110.

065

.36.

76.

736

7.0

200.

0

Arg

entin

aPo

llutio

n M

anag

emen

t19

989.

34.

60.

60.

34.

62.

314

.57.

2

Boliv

iaD

ecen

traliz

atio

n 20

013.

02.

4

Braz

ilBe

lo H

orizo

nte

Met

ropo

litan

19

957.

03.

57.

03.

51.

00.

517

3.0

86.9

8.0

4.0

196.

098

.5

Tran

spor

t

Braz

ilRe

cife

Tra

nspo

rt D

ecen

traliz

atio

n19

954.

12.

119

9.9

99.9

204.

010

2.0

Braz

ilSã

o Pa

ulo

Inte

grat

ed U

rban

19

9884

.038

.87.

95.

01.

21.

293

.145

.0

Tran

spor

t (BF

)

Braz

ilRi

o de

Jan

eiro

Mas

s Tr

ansit

1998

299.

217

9.5

16.8

6.5

316.

018

6.0

Braz

ilSa

lvad

or U

rban

Tra

nspo

rt19

9930

5.5

148.

11.

51.

51.

00.

430

8.0

150.

0

Braz

ilFo

rtale

za U

rban

Tra

nspo

rt20

0117

6.0

85.0

176.

085

.0

Braz

ilSã

o Pa

ulo

Line

420

021,

800.

020

9.0

1,80

0.0

209.

0

Braz

ilRe

cife

Urb

an U

pgra

ding

2002

20.0

11.0

2.0

1.0

22.0

12.0

Col

ombi

aRe

gula

tory

Ref

orm

Tec

hnic

al

1997

2.0

1.0

2.0

1.0

Ass

istan

ce

Col

ombi

aBo

gota

Urb

an S

ervi

ces

2003

43.5

20.0

43.5

20.0

22.3

10.3

27.6

17.7

136.

968

.0

Mex

ico

Med

ium

-Size

Citi

es U

rban

Tra

nspo

rt19

9334

0.0

145.

080

.035

.047

.020

.046

7.0

200.

0

Mex

ico

Seco

nd A

ir Q

ualit

y an

d Tr

ansp

ort

2004

100.

070

.0

Proj

ect

Peru

Lim

a U

rban

Tra

nspo

rt20

0394

.035

.0

Vene

zuel

a,

Urb

an T

rans

port

1993

26.0

9.0

146.

053

.08.

06.

040

.032

.022

0.0

100.

0

Rep.

Bol

. de

Vene

zuel

a,

Car

acas

Slu

m U

pgra

ding

1999

27.0

11.0

3.0

1.0

30.0

12.0

Rep.

Bol

. de

Mid

dle

East

and

N

orth

Afr

ica

Regi

on

Alg

eria

Tran

spor

t Tec

hnic

al A

ssist

ance

2002

3.9

2.8

3.9

2.8

Leba

non

Urb

an T

rans

port

Dev

elop

men

t20

0275

.030

.035

.030

.05.

05.

011

5.0

65.0

Leba

non

Firs

t Mun

icip

al In

frast

ruct

ure

2000

72.8

72.8

5.8

5.8

78.6

78.6

Tuni

siaTr

ansp

ort S

ecto

r Inv

estm

ent

1998

a44

.03.

718

.33.

82.

97.

64.

815

.170

.0

(APL

Pha

se I)

Tuni

siaTr

ansp

ort S

ecto

r Inv

estm

ent

2001

a16

.111

.010

.24.

213

.011

.439

.326

.6

(APL

Pha

se II

)

Sout

h A

sia

Regi

on

Bang

lade

shD

haka

Urb

an T

rans

port

1999

50.1

37.6

52.4

40.2

53.0

40.3

19.9

14.8

5.4

4.0

11.0

6.1

34.4

34.0

226.

217

7.0

Bang

lade

shM

unic

ipal

Ser

vice

s19

991.

01.

09.

08.

010

.09.

0

Indi

aM

umba

i Urb

an T

rans

port

2002

57.0

33.0

25.0

14.5

2.0

1.2

7.0

4.1

26.0

15.0

590.

034

1.5

6.0

3.5

107.

061

.918

.010

.483

8.0

485.

0

Indi

aTa

mil

Nad

u U

rban

Dev

elop

men

t - II

1999

54.9

28.2

54.9

28.2

Tota

ls (a

ll re

gion

s)1,

720.

973

4.9

1,42

8.8

768.

655

5.0

313.

484

.847

.786

.943

.160

9.6

430.

02,5

43.7

498.

31,

639.

796

7.4

59.0

30.0

239.

716

0.0

251.

917

7.8

206.

615

2.51

0,35

4.6

4,86

1.0

Perc

enta

ge16

.615

.113

.815

.85.

46.

40.

81.

00.

80.

95.

98.

824

.610

.315

.819

.90.

60.

62.

33.

32.

43.

72.

03.

110

0.0

100.

0

No

te: H

OV

= h

igh

occ

upan

cy v

ehic

les.

Fig

ures

in t

he s

had

ed a

rea

rep

rese

nt lo

an a

mo

unts

allo

cate

d fo

r a

com

po

nent

.

a. A

dap

tab

le P

rog

ram

Len

din

g, w

here

“le

ndin

g”

refe

rs t

o a

n In

tern

atio

nal B

ank

for

Rec

ons

truc

tion

and

Dev

elo

pm

ent

loan

or

Inte

rnat

iona

l Dev

elo

pm

ent

Ass

oci

atio

n cr

edit.

Sour

ce: W

orld

Ban

k p

roje

ct fi

les.

Note: Background reports for the review (indicated with symbol §) were commissioned with theassistance of national trust funds (France, Japan, the Netherlands, and Spain) or direct governmentfinance (Germany and the United Kingdom). They are all posted on the Urban Transport StrategyReview Web page, and can be downloaded directly from that page, which may be accessed fromthe World Bank Transport Web site at http://www.worldbank.org/transport.

The word processed describes informally reproduced works that may not be commonly availablethrough libraries.

Aeron-Thomas, A., A. J. Downing, G. D. Jacobs, J. P. Fletcher, T. Selby, and D. T. Silcock. 2002.“Review of Road Safety Management Practice: Final Report.” TRL Ltd., with Ross Silcock,Babtie Groups, Ltd.

Affonso S. A., F. Rezende, and F. P. Vitor. 1998. “Peace within Traffic: A Revolution of Attitudes inBrazilia.” In Freeman, P., and C. Jamet, eds., Urban Transport Policy: A SustainableDevelopment Tool. Proceedings of the International Conference CODATU VIII, Cape Town,South Africa. Rotterdam: Balkema.

Amundsen, F. H. 1996. “Review of World Bank Experience in Traffic Safety Concerning Motorizedand Non-Motorized Traffic (1989–94).” TWU. World Bank, Washington, D.C. Processed.

APOYO Group. 1999. “Profile of Socioeconomic levels in Metropolitan Lima.” Apoyo, Opinion yMercadeo.

Aragão, J. J., A. Brasileiro, and R. Marar. 1998. “The Brazilian Bus Industry and the New Legislationon Public Procurement Procedures.” In Freeman, P., and C. Jamet, eds., Urban Transport Policy:A Sustainable Development Tool. Proceedings of the International Conference CODATU VIII,Cape Town. Rotterdam: Balkema.

§ Arcadis Bouw/Infra. 2000. The Development of Logistic Services (The Netherlands).http://www.worldbank.org/transport

Ardila, A., and G. Menckhoff. Forthcoming. “Busways: Lessons from Latin American Cities.” WorldBank, Washington, D.C.

Armstrong-Wright, A. 1986. “Urban Transport Systems: Guidelines for Examining Options.” WorldBank Technical Paper, Urban Series No. 52. Washington, D.C.

———. 1993. Public Transport in Third World Cities. TRL State of the Art Review. London: HMSO.Associacão Nacional de Transportes Públicos. 1999. O Transporte Clandestino no Brazil.

Documentos Setoriais. São Paulo.Barrett, R. 1986. “Institution Building for Traffic Management.” Technical Paper 78. World Bank,

Washington, D.C.Barter, P. 1999. “Transport and Urban Poverty in Asia: A Brief Introduction to the Key Issues.”

Regional Development Dialogue 20 (1).§ BB&J Consult. 2000. “Implementation of Rapid Transit (Spain).” http://www.worldbank.org/transport§ BCEOM. 2000. “Review of French Experience in Institutional Framework for Planning and

Programming in Urban Transport (France).” http://www.worldbank.org/transportBertaud, A. 1999. “Bangalore Land-Use Issues.” World Bank, Washington, D.C.Bertaud, A., and B. Renaud. 1997. “Socialist Cities Without Land Markets.” Journal of Urban

Economics 41.

BIBLIOGRAPHY 193

BIBLIOGRAPHY

194 BIBLIOGRAPHY

Bloy, E. 2001. “Rentabilité et financement des micro-entreprises de transport collectif en Afriquesubsaharienne.” Sub-Saharan African Transport Policy Program Report 54F. World Bank,Washington, D.C.

Burchell, R. W. 1998. The Costs of Sprawl—Revisited. Transit Co-operative Research Program Report39. Transportation Research Board. Washington, D.C.: National Academy Press.

Button, K. J., and E. T. Verhoef, eds. 1998. Road Pricing, Traffic Congestion and the Environment.Cheltenham: Edward Elgar.

§ Caetano Roca Giner (PriceWaterhouseCoopers). 2000. “Conversion of Railway Lines for SuburbanPassenger Services (Spain).” http://www.worldbank.org/transport

Canning D., and E. Bennathan. 2000. “The Social Rate of Return on Infrastructure Investments.”Policy Research Working Paper 2390. World Bank, Washington, D.C.

§ CATRAM. 2000. “Land Use and Transport Development: The Case Study of Cairo, Egypt (France).”http://www.worldbank.org/transport

Centro de Estudios del Transporte del Area Metropolitana. 1999. “Estudio de Transporte yCirculacion Urbana.” Universidad de Buenos Aires.

Cervero, R. 1998. “Paratransit: The Gap Fillers.” Habitat Debate 4 (2). ———. 2001. “Informal Transport: Mobility Options for the Developing World.” Paper Prepared for

UNCHS (Habitat), Nairobi. Conseil National des Transports. 1999. “Les transports et l’environnement: vers un nouvel equili-

bre.” La Documentation Française, Paris. § Cracknell, J. 2000. “Experience in Urban Traffic Management and Demand Management in

Developing Countries (United Kingdom).” http://www.worldbank.org/transport Cusset, J. M. 1998. “Accessibility to Urban Services and Mobility Needs of Peripheral Population:

The Case of Ougadougou.” In Freeman, P., and C. Jamet, eds., Urban Transport Policy: ASustainable Development Tool. Proceedings of the International Conference CODATU VIII,Cape Town. Rotterdam: Balkema.

De Langen, M., and R. Tembele, eds. 2000. “Productive and Liveable Cities: Guidelines forPedestrian and Bicycle Traffic in African Cities.” Interface for Cycling Expertise, Utrecht, theNetherlands.

De St. Laurent, B. 1998. “Overview of Urban Transport in South Africa.” In Freeman, P., and C. Jamet,eds., Urban Transport Policy: A Sustainable Development Tool. Proceedings of the InternationalConference CODATU VIII, Cape Town. Rotterdam: Balkema.

Department of Environment, Transport, and the Regions. 1998. “A New Deal for Transport: Betterfor Everyone.” Cmd 3950. London: HMSO.

Diaz Olvera, L., D. Plat, and P. Pochet. 1998. “Les multiple visages de la mobilité feminine dans lesvilles d’Afrique.” In Freeman, P., and C. Jamet, eds., Urban Transport Policy: A SustainableDevelopment Tool. Proceedings of the International Conference CODATU VIII, Cape Town.Rotterdam: Balkema.

Dimitriou, H. 1998. “Developing a Strategy to Address Problems of Increasing Motorization.”Report prepared for the United Nations Development Programme.

Dollar, D., and A. Kraay. 2001. “Growth is Good for the Poor.” Policy Research Working Paper. WorldBank, Washington, D.C. http://www.econ.worldbank.org

Dourthe, A., M. Wityk, H. Malbran, and O. Figueroa. 1998. “Dereglementation et re-reglementationdu transport public urbain: le Cas de Santiago” In Freeman, P., and C. Jamet, eds., UrbanTransport Policy: A Sustainable Development Tool. Proceedings of the InternationalConference CODATU VIII, Cape Town. Rotterdam: Balkema.

BIBLIOGRAPHY 195

Dowall, D. 1995. “The Land Market Assessment: A New Tool for Urban Management.” UrbanManagement Program Discussion Paper, World Bank, Washington, D.C.

Dutt, A. K., Y. Xie, F. J. Costa, and Z. Yang. 1994. “City Forms of China and India in GlobalPerspective.” In Dutt, A. K., F. J. Costa, A. G. Noble, and S. Aggarwal, eds., The Asian City:Processes of Development, Characteristics and Planning. Boston: Kluwer AcademicPublishers.

ECLAC (Economic Commission for Latin America and the Caribbean). 1991. “The Impacts ofSubsidies, Regulation and Different Forms of Ownership on the Quality and OperationalEfficiency of Urban Bus Services in Latin America.” Santiago, Chile.

Eskeland, G. S., and S. Devarajan. 1996. “Taxing Bads by Taxing Goods: Pollution Control withPresumption Charges.” Directions in Development Series. World Bank, Washington, D.C.

Eskeland, G. S., and J. Xie. 1998. “Integrating Local and Global Environmental Benefits:Mechanisms and Case Studies in the Transport Sector.” World Bank, Washington D.C.

European Commission. 2000. “Developing Urban Management and Safety.” Report of an interna-tional research team (The DUMAS Report). European Union, Brussels, Belgium.

European Conference of Ministers of Transport. 1995. “Urban Travel and SustainableDevelopment.” OECD/ECMT, Paris.

———. 1999a. “Improving the Environmental Performance of Vehicles: Fleet Renewal andScrappage Schemes.” Report of the ECMT Committee of Deputies. Paris.

———. 1999b. “Improving Transport for People with Mobility Handicaps: A Guide to GoodPractice.” Paris.

———. 1999c. “Traffic Congestion in Europe.” Report of 110th Round Table on TransportEconomics. Paris.

Fouchier, V. 1997. “Les densites urbaines et le developpement durable: ler cas de l’Ile de France etdes villes nouvelles.” Secretariat General du Groupe Centrale des Villes Nouvelles, Paris.

Fouracre, P. R., R. J. Allport, and J. M. Thompson. 1990. “The Performance and Impact of Rail MRTin Developing Countries.” TRL Research Report RR278. Transport Research Laboratory,Crowthorne, United Kingdom.

Fouracre, P. R., A. Astrop, and D. Maunder. 1999. “Accessibility for the Urban Poor.” TRL ProjectReport 153/99. Transport Research Laboratory, Crowthorne, United Kingdom.

Frame, G. 1999. “Traffic Management and Road Safety in World Bank Projects in Chinese Cities: AReview.” World Bank, Washington, D.C.

Gallagher, R. 1992. The Rickshaws of Bangladesh. Dhaka, Bangladesh: University Press. Gannon, C., and Zhi Liu. 1997. “Poverty and Transport.” Discussion Paper TWU-30. Transport, Water,

and Urban Development, World Bank, Washington, D.C.Gardner, G. 1993. “The Performance of Light Rapid Transit in Developing Countries.” TRL Project

Report. Transport Research Laboratory, Crowthorne, United Kingdom. Gardner, G., P. R. Cornwell, and J. A. Cracknell. 1991. “The Performance of Busway Transit in

Developing Countries.” TRL Research Report RR329. Transport Research Laboratory,Crowthorne, United Kingdom.

Garreau, J. 1991. Edge City: Life on the New Frontier. New York: Doubleday.Ghee, C., D. Silcock, A. Astrop, and G. D. Jacobs. 1997. “Socio-economic Aspects of Road

Accidents in Developing Countries.” TRL Report 247. Transport Research Laboratory,Crowthorne, United Kingdom.

§ Godard, X., and L. Diaz Olvera. 2000. “Poverty and Urban Transport: French Experience andDeveloping Cities (France).” http://www.worldbank.org/transport

196 BIBLIOGRAPHY

Gold, P. 1999. Using Engineering to Reduce Accidents. Inter-American Development Bank,Washington, D.C.

Gomez, L. M. 2000. “Gender Analysis of Two Components of the World Bank Transport Projects inLima, Peru: Bikepaths and Busways.” World Bank Internal Report, World Bank, Washington,D.C.

§ GTZ (Deutsche Gesellschaft für Technische Zusammenarbeit). 2002. “Urban Transport StrategyReview: Experiences from Germany and Zurich (Germany).” http://www.worldbank.org/trans-port

Gwilliam, K. M. 1996. “Getting the Prices Wrong: A Tale of Two Islands.” Infrastructure Note UT-6.Transport, Water, and Urban Development, World Bank, Washington, D.C.

———. 2000a. “Natural Gas as a Transport Fuel.” Infrastructure Note UT-8. Transport, Water, andUrban Development, World Bank, Washington, D.C.

———. 2000b. “Private Participation in Public Transport in the Former Soviet Union.” DiscussionPaper TWU-40. Transport, Water, and Urban Development, World Bank, Washington, D.C.

———. 2001. “Competition in Passenger Transport in the Developing World.” Journal of TransportEconomics and Policy 35 (1/January).

Gwilliam, K. M., and R. G. Scurfield. 1996. “Constructing a Competitive Environment in Public RoadPassenger Transport.” Discussion Paper TWU-24. Transport, Water, and Urban Development,World Bank, Washington, D.C.

Gwilliam, K. M., A. J. Kumar, and R. T. Meakin. 2000. “Designing Competition in Urban BusTransport: Lessons from Uzbekistan.” Discussion Paper TWU-41. Transport, Water and UrbanDevelopment, World Bank, Washington, D.C.

§ Halcrow Fox, in association with Traffic and Transport Consultants. 2000a. “Mass Rapid Transit inDeveloping Countries (United Kingdom).” http://www.worldbank.org/transport

§ Halcrow Fox. 2000b. “Review of Urban Public Transport Competition (United Kingdom).”http://www.worldbank.org/transport

Hau, T. 1992a. “Economic Fundamentals of Road Pricing: A Diagrammatic Analysis.” PolicyResearch Working Paper 1070. Transport, Water, and Urban Development, World Bank,Washington, D.C.

———. 1992b. “Congestion Charging Mechanisms for Roads: An Evaluation of Current Practice.”Policy Research Working Paper 1071. Transport, Water, and Urban Development, World Bank,Washington, D.C.

Heggie, I. G., and P. Vickers. 1998. “Commercial Management and Financing of Roads.” World BankTechnical Paper 409, World Bank, Washington, D.C.

Henry, E. 2000. “Mobility, Motorization and Income: Does Urban Growth Mean More Transport?”The French Research Institute for Development (IRD—Institut de recherché pour le développe-ment). henry@bondy.ird.fr.

Hidalgo Guerrero, D. 2001. “TransMilenio: The Mass Transport System of Bogotá.” A paper pre-pared for the Latin American Urban Public Transport Congress (KLATPU), Havana.

Holgate, S. T., J. M. Samet, H. S. Koren, and R. L. Maynard, eds. 1999. Air Pollution and Health.London: Academic Press.

§ Howe J., and D. Bryceson. 2000. “Poverty and Urban Transport in East Africa (Netherlands).”http://www.worldbank.org/transport

Hughes, G., and M. Lovei. 1999. “Economic Reform and Environmental Performance in TransitionEconomies.” Technical Paper 446. World Bank, Washington, D.C.

Hwang, K. Y., B. Son, and J. K. Eom. 1999. “Effect of Congestion Pricing at the Namsan Tunnels inSeoul.” Journal of the Eastern Asia Society for Transportation Studies 3 (4).

BIBLIOGRAPHY 197

Ingram, G., and Zhi Liu. 1999. “Determinants of Motorization and Road Provision.” Policy ResearchWorking Paper 2042. World Bank, Washington, D.C.

Interface for Cycling Expertise. 1997. Cycling Promotion and Bicycle Theft. Local background stud-ies in León, Nicaragua; Lima, Peru; New Delhi, India; Guangzhou, China; and Accra, Ghana.Utrecht, Netherlands.

§ ———. 2000. “Nonmotorized Traffic in Developing Countries (Netherlands).” http://www.world-bank.org/transport

IPCC (Intergovernmental Panel on Climate Change). 1996. Climate Change 1995: The Economicand Social Dimensions of Climate Change. Cambridge (U.K.): Cambridge University Press.

Jacobs, G., and A. Aeron-Thomas. 2000. “A Review of Global Road Accident Fatalities.” Paper com-missioned by the Department for International Development (United Kingdom) for the GlobalRoad Safety Partnership.

JICA-Chodai. 1996. “Master Transportation Plan of Bogotá.” Bogotá.Jones, P. 1998. “Urban Road Pricing: Public Acceptability and Barriers to Implementation.” In Road

Pricing, Traffic Congestion and the Environment Issues of Efficiency and Social Feasibility.Cheltenham: Edward Elgar.

Joos, E. 2000. “Kunming: A Model City for a Sustainable Development and Transport Policy inChina.” Public Transport International May 24–27.

Karan, P. P. 1994. “The Distribution of City Sizes in Asia.” In Dutt, A. K., F. J. Costa, S. Aggarwal, andA. G. Noble, eds., The Asian City; Processes of Development, Characteristics and Planning.Boston: Kluwer Academic Publishers.

Kojima, M., C. Brandon, and J. Shah. 2000. “Improving Urban Air Quality in South Asia by ReducingEmissions from Two-Stroke Engine Vehicles.” South Asia Environment Unit, World Bank,Washington, D.C.

Kojima, M., and M. Lovei. 2001. “Urban Air Quality Management: Coordinating Transport,Environment, and Energy Policies in Developing Countries.” World Bank Technical Paper 508,World Bank, Washington, D.C.

Koster, J. H., and M. de Langen. 1998. “Preventive Transport Strategies for Secondary Cities.” InFreeman, P., and C. Jamet, eds., Urban Transport Policy: A Sustainable Development Tool.Proceedings of the International Conference CODATU VIII, Cape Town. Rotterdam:Balkema.

———, eds. 2001. Low-Cost Mobility in African Cities. Proceedings of the Expert Group Meeting,Delft, Netherlands. Delft: IHE.

Kranton, R. 1991. “Transport and the Mobility Needs of the Urban Poor: An Exploratory Study.”Discussion Paper INU 8. Infrastructure and Urban Development Department, World Bank,Washington, D.C.

Krugman, P. R. 1991. Geography and Trade. Belgium: Leuven University Press.Kumarage, A. S. 1998. Formulation of a Policy Framework for Poverty Alleviation: Transport.

University of Moratuwa, Sri Lanka.Kwakye, E., P. Fouracre, and D. Osufa-Dorte. “Developing Strategies to Meet the Transport Needs

of the Poor in Ghana.” World Transport Policy and Practice 3.Lee, D. B. 1999. “The Efficient City: Impacts of Transport Pricing on Urban Form.” Paper presented

at the Annual Conference of the Associated Schools of Collegiate Planning, Cambridge, Mass. Lee, K. S. 1989. The Location of Jobs in a Developing Metropolis. A World Bank Research

Publication. New York: Oxford University Press. Litman, T. 1999. “Distance-Based Charges: A Practical Strategy for More Optimum Vehicle Pricing.”

Transportation Research Board’s 78th Annual Meeting, Washington, D.C.

198 BIBLIOGRAPHY

§ Louis Berger, s. a. 2000. “Review of French Experience With Respect to Public Sector Financing ofUrban Transport (France).” http://www.worldbank.org/transport

Lovei, M. 1996. “Phasing Out Lead from Gasoline: Worldwide Experience and Policy Implications.”Environment Department Paper 40. World Bank, Washington, D.C.

Lvovsky, K., G. Hughes, D. Maddison, B. Ostro, and D. Pearce. 2000. “Environmental Costs of FossilFuels: A Rapid Assessment Method with Application to Six Cities.” Environment DepartmentPaper 78. World Bank, Washington, D.C.

Madras Metropolitan Development Authority. 1990. “Survey of Pavement Dwellers, Madras City1989–90.” Madras.

Madrid Transport Consortium. 1999. Annual Report. Madrid.Menckhoff, G., and C. Zegras. 1999. “Experiences and Issues in Urban Transport Infrastructure.” IRF

Symposium on Innovative Financing in Transportation Projects, Hanoi.Merilainen, A., and R. Helaakoski. 2001. “Transport, Poverty and Disability in Developing Countries.”

World Bank, Washington, D.C. Draft.Midgley, P. 1994. “Urban Transport in Asia: An Operational Agenda for the 1990’s.” Technical Paper

224. World Bank, Washington, D.C.Mitric, S. 1997. “Approaching Metros as Development Projects.” TWU Discussion Paper 27.

Transport, Water, and Urban Development, World Bank, Washington, D.C.Mohring, H. 1999. “Congestion.” In Essays in Transportation Economics and Policy: A Handbook in

Honor of John R. Meyer. Washington, D.C: Brookings Institution Press. Moser, C. A. 1996. “Confronting Crisis: A Summary of Household Responses to Poverty and

Vulnerability in Four Poor Urban Communities.” Environmentally and Socially SustainableDevelopment Studies and Monographs Series No. 7. World Bank, Washington, D.C.

Moser, C. A., and L. Peake. 1987. Women, Human Settlements, and Housing. London: Tavistock. Narayan, D. 2000a. Voices of the Poor: Can Anyone Hear Us? New York: Oxford University Press.———. 2000b. Voices of the Poor: Crying Out for Change. New York: Oxford University Press.§ NEA Transport Research and Training. 2000. “Relationship between Urban Land Use Planning,

Land Markets, Transport Provisions and the Welfare of the Poor (Netherlands).”http://www.worldbank.org/transport

Newbery, D. 1990. “Pricing and Congestion: Economic Principles Relevant to Pricing Roads.”Oxford Review of Economic Policy 6 (2).

Newman, P. and J. Kenworthy. 1989. Cities and Automobile Dependence: A Sourcebook. Aldershot:Gower Technical.

———. 1999. Sustainability and Cities: Overcoming Automobile Dependence. Washington, D.C:Island Press.

§ PADECO Co. Ltd.. 2000. “Study on Urban Transport Development: Japanese Experience (Japan).”http://www.worldbank.org/transport

Pattison, T., ed. 1999. Jane’s Urban Transport Systems 1999–2000. 18th edition. Coulsdon, UnitedKingdom, and Alexandria, Va.: Jane’s Information Group.

Pendall, R. 1999. “Do Land-Use Controls Cause Sprawl?” Environment and Planning B: Planning andDesign 26.

§ Prointec Inocsa Stereocarto. 2000. “Urban Public Transport: Integration and Financing (Spain).”http://www.worldbank.org/transport

Quinn, D.J. 1998. “The Leeds Guided Busway: A Low Cost Solution for Developing Countries?”In Freeman, P., and C. Jamet, eds., Urban Transport Policy: A Sustainable Development Tool.Proceedings of the International Conference CODATU VIII, Cape Town. Rotterdam:Balkema.

BIBLIOGRAPHY 199

Rebelo, J. 1996. “Essentials for Sustainable Urban Transport in Brazil’s Large Metropolitan Areas.”WPS 1633. World Bank, Washington, D.C.

———. 1999a. “Rail and Subway Concessions in Rio de Janeiro: Designing Contracts and BiddingProcesses.” Viewpoint #183. World Bank, Washington, D.C.

———. 1999b. “Reforming the Urban Transport Sector in the Rio de Janeiro Metropolitan Region:A Case Study in Concessions.” WPS 2096. World Bank, Washington, D.C.

Rebelo, J., and P. Machado. 2000. “The São Mateus–Jabaquara Trolleybusway Concession in Brazil.”Policy Research Working Paper 2353. World Bank, Washington, D.C.

Richardson, H., and C. Bae. 1998. “The Equity Impacts of Road Congestion Pricing.” In Road Pricing,Traffic Congestion and the Environment Issues of Efficiency and Social Feasibility. Cheltenham:Edward Elgar.

Rickert, T. 1998. Mobility for All: Accessible Transportation Around the World. San Francisco: AccessExchange International.

Robinson, R., U. Danielsson, and M. S. Snaith. 1998. Road Management: Concepts and Systems.London: Macmillan.

Ross, A. 1993. “Review of World Bank Experience in Road Safety.” Discussion Paper INU 93.Infrastructure and Urban Development Department, World Bank, Washington, D.C.

Roth, G. 1996. Roads in a Market Economy. Aldershot: Avebury. Sachdeva, Y. P. 1998. “Walk and Bicycle Travel Characteristics in Indian Cities.” In Freeman, P., and

C. Jamet, eds., Urban Transport Policy: A Sustainable Development Tool. Proceedings of theInternational Conference CODATU VIII, Cape Town. Rotterdam: Balkema.

SACTRA (Standing Advisory Committee on Trunk Road Assessment). 1994. “Trunk Roads and theGeneration of Traffic.” United Kingdom.

Schipper, L., and C. Marie-Lilliu. 1999. “Transportation and CO2 Emissions: Flexing the Link—A Pathfor the World Bank.” Environment Papers No 69. Environmentally and Socially SustainableDevelopment, World Bank, Washington, D.C.

Schwela, D., and O. Zali. 1999. Urban Traffic Pollution. London: E & FN Spon. Small, K. 1992. Urban Transportation Economics. Chur: Harwood Academic Publishers. § Sodeteg. 2000. “Review of French Experience in Respect of Characteristics and Potentialities of Urban

Fixed Track Systems in Developing Countries (France).” http://www.worldbank.org/transport§ Systra. 2000. “Review of French Experience in Private Financing of Public Urban Transport

(France).” http://www.worldbank.org/transport Teurnier, P., and B. Mandon-Adolehoume. 1994. “L’Integration du Transport Artisanal dans un

Service Public de Transport Urbain: Le Cas de Dakar. ” Paris: TTD.Thompson, I. 1993. “Improving Urban Transport for the Poor.” CEPAL Review (49).TRB (Transportation Research Board). 1998a. “Guidebook for Evaluating, Selecting and

Implementing Fuel Choices for Transit Bus Operations.” Transportation Co-operative ResearchProgram Report 38. TRB, Washington, D.C.

———. 1998b. “The Costs of Sprawl Revisited.” Transportation Co-operative Research ProgramReport 39. TRB, Washington, D.C.

TRL (Transport Research Laboratory). 1985. “The Demand for Bus Transport.” TRL, Crowthorne,United Kingdom.

———. 1995. “Costing Road Accidents in Developing Countries.” Overseas Road Note ORN 10.TRL, Crowthorne, United Kingdom.

§ TRL (Transport Research Laboratory) and R. Silcock. 2000. “Review of Road Safety in Urban Areas(United Kingdom).” http://www.worldbank.org/transport

United Nations. 1996. World Urbanization Prospects. New York.

200 BIBLIOGRAPHY

UNCHS (United Nations Centre for Human Settlement). 1998. “Global Urban Indicators Database.”Urban Indicators Programme, Nairobi.

Verhoef, H., P. Nijkamp, and P. Rietveld. 1995. “The Economics of Regulatory Parking Policies: The(Im)possibilities of Parking Policies in Traffic Regulation.” In Transportation Research, Part A,Policy and Practice 29 (A): 2.

Weaver, C. S., and L. M. Chan. 1996. “Bangkok Particulate Study.” Report to the Thai Governmentand the World Bank. World Bank, Washington, D.C.

Wheaton, W. 1996. Land Use and Density in Cities with Congestion. Cambridge, Mass.: MIT Centerfor Real Estate.

White, P.R. 1999. “Impacts of the Cairo Metro.” Paper presented at the Sixth InternationalConference on Competition and Ownership in Land Passenger Transport, Cape Town.

Willoughby, C. 2000a. “Managing Motorization.” TWU Discussion Paper 42. Transport, Water, andUrban Development, World Bank, Washington, D.C.

———. 2000b. “Singapore’s Experience in Managing Motorization and its Relevance to OtherCountries.” TWU Discussion Paper 43. Transport, Water, and Urban Development, World Bank,Washington, D.C., and the Web site of the Singapore Land Transport Authorityhttp://www.lta.gov.sg.

World Bank. 1975. “Urban Transport.” A World Bank Policy Paper. Washington, D.C.———. 1986. “Urban Transport: Sector Policy Paper.” A World Bank Policy Paper. Washington, D.C.———. 1995. Bureaucrats in Business: The Economics and Politics of Government Ownership.

Oxford: Oxford University Press. ———. 1996. “Sustainable Transport: Priorities for Policy Reform.” A World Bank Policy Paper.

Washington, D.C.———. 1998. “Fuel for Thought: a New Environmental Strategy for the Energy Sector.” A World

Bank Policy Paper. Washington, D.C.———. 1999. “Bangkok Urban Transport Sector Review.” Washington, D.C.———. 2000a. “Cities in Transition: World Bank Urban and Local Government Strategy.” World

Bank Policy Paper. Washington, D.C.———. 2000b. “Reforming Public Institutions and Strengthening Governance.” Poverty Reduction

and Economic Management Network, Washington, D.C.———. 2000c. World Development Report. Washington, D.C.———. 2001. “Mexico Energy Environment Review.” ESMAP (Energy Sector Management

Assistance Programme). Report 241/01. Washington D.C. WHO (World Health Organization). 1992. Urban Air Pollution in Megacities of the World. Oxford

U.K. and Cambridge, Mass.: Blackwell Reference.———. 1999. World Health Report. Geneva.Wright, C.L., ed. 2001. Facilitando o Transporte para Todos. Banco Interamericano de

Desenvolvimento, Washington D.C. Zegras, C. 1998. “The Costs of Transportation in Santiago de Chile: Analysis and Policy

Implications.” Transport Policy Journal 5.

Accessibility poor, 27Accidents, incidence, 65, 68; location, 68; pre-

vention programs, 68–69; Sri Lanka, 66; sta-tistics, 68

Adjudication, land invasions and informalacquisitions, 19

Advisory services, 179Affordability, 34–36Agglomeration economies, 5, 8, 10, 16Air pollution, 42–59; economics, 8; mortality

and morbidity and, 39; pollutants, 42–45;see also Environment

Ancillary infrastructure, costs, 19Area licenses, 137Argentina, safety and security, 66–67, 74Authority: contiguous, 155; overlapping, 155,

160–61; strategy, 167–68

Bangkok Transit System (BTS), 15Basic knowledge, 60Blocking, 96Borrowing, 146Bottlenecks, 87Brazil, 14, 72, 90, 116Buenos Aires Urban Transport Project, 66–67Built environment, 39Buses, 58, 80–81; accidents, 68; design, 112;

financing, 174Bus sector, 93–106; competition, 95–96; con-

tracts, 98–100, 105; efficiency, 94–95;financing, 94; franchises, 100, 105; institu-tional requirements, 98; mass transit, 178;monopolies, 93; ownership, 93–94; policy,98–101; private sector role, 15; rapid transitsystem, 119; ridership, 93; Uzbekistan, 100

Busways, 112; financing, 118

Capability, institutions, 167Capacity building, 179; institutions, 163–64, 165Capacity extensions, economics, 87–88Capital cities, 16, 22Car ownership, income and, 8Carbon dioxide (CO2), 40

Carbon monoxide (CO), 43, 45Catalyst technologies, 51Catalytic converters, 50, 51Charges: system, 140–41; vehicle use, 84China, 93, 128, 131City characteristics, 6–7City development strategies (CDS), 5, 11–12,

22–23Civil society, engaging, 164–65Collaboration, 181–82; voluntary, 160Collusion, prevention, 105Colombia, 74, 119, 129Community severance, 42Commuters, 25Competition, 34, 37, 171; bus sector, 95–96;

effective, 98; informal sector, 103, 104;managing, 96, 98; strategy, 107

Comprehensive Development Framework, 170,181–82

Compressed natural gas (CNG), 47–48Concessioning, 95–96, 173Conflict reduction, institutions, 160 Congestion: city development, 7, 10, 22; envi-

ronment, 58; impacts, 142–43; povertyreduction and, 29, 32–33; reduction, 110

Congestion pricing, 135, 136–38Contracts, 15, 98–100; gross cost service con-

tracting, 95; net cost service contracting, 95Control system, 19Coordinating authority, 157Coordination, 22, 116, 143–44; committee,

156–57; functional, 168; institutions,154–56, 160–62, 163–64, 167–68; jurisdic-tional, 167–68

Cordon pricing, 137Cost-benefit, capacity extensions, 88Costs, 37; accidents, 65; ancillary infrastructure,

19; congestion, 142–43; cycling, 133–34;demand management, 83; I/M programsand, 54–55; infrastructure, 19, 20; MRT, 112;metros, 115–17; recovering, 143; unit trans-port, 19; undercharging, 19; sprawl, 19; seealso Economics; Finance

INDEX 201

INDEX

202 INDEX

Counterpart funding, 148Country Assistance Strategy, 181Credit, bus sector, 106Crime, 73, 75Cultural factors, 28Curb rights, 105Cycling, 30–31, 125, 126–28; attitudes, 127;

women, 127Cyclists, 126–27, 131

Decentralization, 14, 15, 22; facilitating, 171–73;policy, 172; political and financial responsi-bility, 1–2; progress, 172; WB agenda,172–73

Demand for transport, 12Demand management, 60, 83–84, 91, 171,

177–78 Densification, 17Development: challenges, 169–70; control

skills, 23; planning, 14Disabilities, 27Distance, 7, 26–27District interests, 155“Doing the right thing,” 141–42Drunken driving, preventing, 71

Easements, 13, 111Economics: capacity extensions, 87–88; cities

in developing countries, 5; deconcentra-tion, 16; dynamics, 8, 9–11, 22; efficiency,28; paratransit, 104; poverty impact ofurban roads, 89; subsidies, 142; see alsoAgglomeration economies; Costs; Finance

Edge cities, 11Education sector, 21Efficiency: development strategies for, 12–15;

economists and, 16; equity vs., 31–32Electric vehicles, 48–49Electronic road pricing (ERP), 137–38 Emergency medical services, 71Emissions, 40, 45, 50–51; high emitters, 53Energy-poverty, 27Engineering, institutions, 165–66Engines, 51–52; two-stroke, 51–52Environment, 39–63; densification and, 17; dump-

ing, 56–57; MRT and, 111–12; policy, 177;progress, 177–78; protecting, 176–79; stan-

dards application, 53–58; strategies, 60–61;WB agenda, 178–79; see also air pollution

Equity, 144–46Ethanol, 48Ethiopia, road safety, 67, 72Exclusion, 28

Facilities location, 21Fares: affordable, 34–36, 37, 174; discrimina-

tion, 145; exemptions, 36, 145; informalsector, 104; levels, 144; reduced, 21, 145;service quality and, 181; structure, 145

Fatalities, causes, 70, 71Finance, 37, 146–50, 172; buses, 94, 174; chan-

nels, 180; cross-modal transfers, 122; infra-structure, 146–47, 150–51; integrated,149–50; metros, 117; MRT, 117–20, 122–23;NMT, 133–34; planning, 22; private, 15, 88,117, 147–48; public transport, 141–46; roadmaintenance, 77–79; strategy, 151; see alsoCosts; Economics

Firms, location, 10–11Focus on poverty: policy, 170; progress, 170-71;

strengthening, 170–71; WB agenda, 171Framework, urban planning and market forces, 16Franchises, 173–74; bus sector, 95, 100, 105–6Freight transport policies, 13–14 Fuel, 40–41; adulteration, 46; biofuels, 48; con-

sumption, 49–50; diesel, 41, 55–56; econo-my, 50; fossil, 41, 43, 44; improving quality,45–47; gasoline octane, 50, 52; leaded vs.unleaded gasoline, 51; policy, 45–50; regu-lation, 46; staged strategy, 51; standards,50; substitutes, 47–49; tax, 139–40

Fuel-cell technology, 50Full integration, 131Full segregation, 131Funding, 148–50

Gasoline. See FuelsGender-related disadvantages, 27–28, 127Global Environment Fund (GEF), 41Global overlay, 41Global Road Safety Partnership, 67Global warming, 40–42Government: budget restructuring, 164; freight

movement, 14

INDEX 203

Greenhouse gases (GHG), 40–42; technologicalmeasures, 41

Grid networks, 87Gross domestic product (GDP), 8Growth, urban roads and the poor, 89Guidelines, road safety, 69

Hanging back, 96Health services, 21Hierarchy, conflict reduction, 160Hopewell Project, 120Household: budget, 26; characteristics, 28Housing, 19Hoy no circula {do not drive today] scheme, 59Human resource development, 167, 172; see

also StaffHybrid diesel-electric vehicles, 49Hydrogen fuel cell, 49, 50

Implementation and enforcement, institutions,162–63

Incentive systems, 60Income, 6, 7, 10; car ownership and, 8; car-

users, 32; personal vs. national, 28–29; pro-portion spent on transport, 26; urban roadsand, 89–90

Income poor, 27, 28Independent authority, 157India, 6, 72, 165Industrial development, 14Industrialized countries, urban sprawl, 11Informal sector services, 35Informal transport sector, 37; congestion, 104;

controlling operating practices, 103–4;environmental impact, 104; formal servic-es and, 104; future, 104–6; negativeimage, 103; paratransit, 101–4; problems,102–3

Infrastructure, 22, 28, 37, 84–88; charges,135–36, 146–47; concessioning, 173; costs,20; design, 69, 175–76; finance, 150–51;NMT policies, 129–31; policy focus, 29–32;public financing, 146–47; roads, 177

Inspection and maintenance (I/M) programs,53–55; introduction, 54

Institutions: capability, 167; capacity building,165; civil society, engaging, 164–65; con-

flict reduction, 160; coordination, 154–56,160–62, 163–64, 167–68; development,172; environment, 60–61; functional basis,156–59; human resource development,167; implementation and enforcement,162–63; jurisdiction, 167–68; NMT, 132–34;organization, 156–65; planning and engi-neering, 165–66; planning and operations,22, 161–62; police role, 162–63; public andprivate sectors, coordinating, 163–64; safe-ty, 71–72, 176; skills, 166–67; strategy,167–68; strengthening, 153–68; traffic man-agement and enforcement, 166; training,167; weaknesses, 154

Instruments, 179–80Intelligent Transport Systems (ITS), 71Intergovernmental transfers, 148–49 Intersections, 131Intrasectoral coordination, 149

Japan, 20Jurisdiction, institutions, 155, 167–68

Knowledge building, 179Kyrgyz Republic, 78

Land: characteristics, 17–18; developmentcosts, 19; readjustment, 20; values, 21

Land use: administrative control, 17–18; change,19; congestion, 143; control, 14; coordina-tion, 155; implementation, 18–19; marketsand distribution, 19–21; MRT and, 111; plan-ning, 15–16, 18; planning and management,16–19; roads, 13; technical skills, 18

Lead, 42–43, 44, 45Leapfrogging, 11Legal framework, 160; see also RegulationsLending: to cities, 180–81; high-impact,

179–80; programmatic, 179Licensing fees, 79Light rapid transit (LRT), 114Limited use, 84Liquefied petroleum gas (LPG), 48Logistic networks, 14

Malaysia, 68Management contracting, bus sector, 95

204 INDEX

Market, 22; activity and, 16; bus sector, 95–96;failure, 96; land-use distribution, 19–21;public transport, 121; structure, 176

Mass rapid transit (MRT), 31, 109–24, 178; ben-efits, structural, 110; environment and,111–12; financing, 110–11, 117–21, 122–23;management, 123; metros, 31, 115–17;objectives and role, 110–11; ownership,117–21; planning, 122; pricing, 121–22,123; project preparation, 120–21; publictransport integration, 121; ridership, 109;strategic objectives, 109–110; strategy,122–23; technology, choice of, 112–17;timeframe, 121; urban structure and landuse and, 111

Mechanized transport, private, 27Medical policies, 71Megacities, 5, 6, 16Merit good, 33Methane, 40, 41Methanol, 48Metros, 31, 115; costs, 115–17; see also Mass

rapid transitMexico, 54, 66, 67Microcredit, 180–81Migration, discouraging, 16Modes of movement, 5–6Monopolies, 93, 105Motor vehicles, ownership and use, 7Motorcycles, 51–52; information base, 52, 53;

injuries, 67–68; standards, 52; taxes andfees, 53; two-stroke engines, 52, 53

Motorized transport, 39; two-wheelers, 6Multimodal integration, 37Multimodal systems, 32Municipal interests, 155

Nepal, 49Netherlands, 17Network: completeness and integrity, 33; plan-

ning, 129Nitrous oxides (NOx), 40, 41, 43Noise, 42Nonmotorized transport (NMT), 30–31, 37, 59,

89, 125–34; financing, 133–34; institutionsand organization, 132–34; policy, 128–32;safety education and training, 132; stake-

holders, 132–33; strategy, 134; traffic man-agement, 131–32; volume, 131

Odds and evens policy, 84Opportunity cost, 41Organization, institutions, 156–65Ownership: MRT, 120–21; procedures, 181Ozone, 43

Pakistan, 74Paratransit, 101–4; future, 104–6; growth, 106;

services provided, 101; strategy, 107Parking: charges, 140; demand management,

83; private nonresidential, 83Partial segregation, 131Participation, 181Passenger transport, 93–108Paving programs, 181Pedestrians, 42, 125, 126; safety and security, 67, 74Peripheral infrastructure, 14Performance, deteriorating, 6, 8Personal characteristics, deprivation and, 27PHARE program, 67Physical disabilities, aids, 27Planning, 174; institutions, 154, 161–62; MRT,

121, 122; strategy, 106–7, 161–62Police: female, 74; officers, 74; role, 162–63;

training, 166Policy: accident prevention, 68–69; bus sector,

98–101; coordination, 21–22, 23; decentral-ization, 172; environmental protection, 177;focus on poverty, 170; fuel, 45–50; fuel taxa-tion, 55–56; infrastructure, 29–32; medical,71; NMT, 128–32; private participation, 173;public expenditures, 57–58; reform, 146;safety and security, 175; social sector, 21–22;system management, 58–59; vehicle, 50–53

Politics, 149; history, 6; processes, 164–65; vio-lence, 74

The Poor, residence, 89Population: concentration, 8; growth, 5; rates, 7Poverty, concept, 25Poverty Reduction Strategy, 2–3; Papers (PRSPs),

170, 181; urban transport and, 25–38Price distortions, 23Prices and pricing, 84, 135–46; congestion,

136–138; efficient, 19; fuel tax, 139–40;

INDEX 205

infrastructure charges, 135–36; MRT,121–22, 123; public transport, 141–46; role,135; strategy, 107, 150–51, 177–78

Primacy, 16Priorities, private sector, 15Private sector: financing, 15, 147–48; institu-

tions, 163–64; MRT, 117–18; participation,mobilizing, 173–75; policy, 173; progress,173-74; public sector and, 156, 163–64;role, 14–15; traffic management, 81–; vehi-cle inspection and maintenance, 54; WBagenda, 174–75

Privatization, 34Property rights, 19Propoor economic growth, 28–29Prototype Carbon Fund, 41Publicly owned lands, recycling, 19Public opinion, 149Public-private participation, 174Public processes, 165Public sector: administrative activities, 21–22;

bus sector, 93–94, 100; expenditure poli-cies, 57–58; infrastructure financing,146–47; pricing, 20; private sector vs., 100

Public transport, 7; accidents, 68; buses, 93;competition, 107; economic liberalization,58; efficiency, 170; implementation, 81–82;markets, 121; MRT integration, 121; pover-ty reduction, 25, 28, 32–34, 37; planningand integration, 106–7; pricing, 107,141–46; priorities, 58; strategy, 106–7; sub-sidies, 57; traffic management, 80–82; seealso Paratransit

Racing, 96Rail systems, 20; investments, 31–32; mass tran-

sit, 178; private sector role, 15; suburban,114–15, 117

Ramsey pricing rules, 143Redevelopment, 19Redistribution, perverse, 21Refineries, 46–47Reform, phased, 100–101Regulations: development strategies, 19–20;

environment, 176; fuel quality, 46; institu-tions, 164; legal framework, 160; municipal,8; traffic law enforcement, 72

Rent-seeking behavior, 148–49Research octane number (RON), 50, 52Resettlement: civil society, 165; design and

implementation, 181; rail systems and, 32Residence: choices, 10, 11; location, 26–27, 28,

89Ridership: buses, 93; MRT, 109 Rights-of-way, 13, 111Ring and radial design, 85, 87Roads: design, 69; infrastructure, 8, 12–13,

150–51, 177; investments, 23, 29–30; locat-ing, 90; low-income areas, 90; mainte-nance, 77–79, 91; network, 85; planning, 91

Road systems, 77–91; capacity extensions,87–88; hierarchy and shape, 85, 87; pover-ty focus, 88–90; private financing, 88; strat-egy, 90–91

Safety, 8, 66–72, 75–76, 79; cost-effectiveness,69; education and training, NMT, 132; evi-dence, 67; funding, 71–72; infrastructuredesign, 69; institutions, 71–72; NMT, 132;policy formation, 68–69; strategies, 75; traf-fic management, 70–71

Safety and security: increasing, 175–76; policy,175, 181; progress, 175-76; WB agenda, 176

Safety-net approach, 33–34Safety poor, 27Scrappage programs, 55Second generation road funds, 78, 79Security, 65–66, 72–76; paratransit, 105; strate-

gies, 75–76; threats to, 73Severance, 42Sexual harassment, 74Signals, 80Singapore, electronic road pricing, 138Size and size distribution, 6Social exclusion, urban poverty and, 25Social sector, 21–22Social security, 21Social violence, 74South Africa, 74, 130South Korea, 137SOx. See Sulfur oxidesSpeed, 7Speed limits, 70Sri Lanka, 66

206 INDEX

Staff, 165–66, 166–67Stakeholders, NMT, 132–33Statistics, road accidents, 68Strategies, 3, 6; city development, 22–23; coor-

dinating sector policies, 21–22; environ-ment, 60–61; institutions, 167–68; motorcy-cles, 52–53; MRT, 122–23; NMT, 134; oper-ational levels, 153–54; poverty focus, 37;pricing and finance, 151; public transport,106–7; roads, 77, 90–91, 150–51; safety andsecurity, 75–76; structural change, 15–21;transport efficiency, 12–15

Structural change: planning capabilities, 22;strategies, 15–21

Subsidies, 57–58, 122, 142, 149Suburban railways, 114–15, 117; see also Rail

systemsSulfur, 51Sulfur oxides (SOx), 43, 45–46Supply efficiency, 141–43; congestion, 142Sustainability, 146, 170–71Synergy, local and global effects, 41System management, policies, 58–59

Tama Garden City Development Project, 20Tanzania, 133 Targeting: congestion, 142; financing, 36–37;

subsidies, 34Taxation, 20; environment, 41–42, 177–78; financ-

ing, 147; fuels, 55–56, 139–40; infrastructure,136–37; relative levels, 56; vehicle, 140

Technology: priorities, 60; supporting develop-ment, 177

Thailand, 15, 46, 120Theft: by stealth, 73; by force, 73Three-wheel motor vehicles, road accidents,

67–68Ticketing systems, 32Time-dependent tolling, 137Time poor, 27Title, 19Tokyu Railway Company, 20Tolls: roads, 88; shadow systems, 15Trade liberalization, 56–57Traffic, 43; authority, 162–63; calming measures,

70; control, 79–80; law enforcement, 72,156, 166; restraint, 58–59, 84

Traffic management, 23, 58, 79–82, 91; agency,70; coordination, 156; cycling and, 131–32;environment, 177; implementation, 81–82;institutions, 166; process, 81; safety andsecurity, 70–71

Training, 166, 167Transport: investments, land values and, 21;

modes, 155–56, 170; patterns, urban poor,25–28; planning, 15–16

Transport sector, growth and change, 45Travel time, 7–8, 26Trickle down, 29Trip rates, 25–26, 27Trucks, road accidents, 68Turning back, 96Two-tier authority, 157

Urban development, 5–23; lower-order, 16;MRT and, 111; planners, 16; strategy, 2–3

Urbanization, 5Urban sprawl, industrialized countries, 11Uzbekistan, 100

Vale-transporte, 147Vandalism, 73–74Vehicles: age limit, 57; charges for use, 84;

electric, 48–49; existing stock, 53; hybriddiesel-electric, 49; policy, 50–53; quality,46; replacement, 41–42, 52, 53, 55; taxes,140; technology, 50–51

Venezuela, 74Versement transport, 147, 160Vietnam, 72Violence, 73Vulnerable populations, protecting, 171

Walking. See Nonmotorized transport;Pedestrians

Women, 27–28; cycling, 127World Bank: contribution, 169–82; policy shifts,

179–80; review content, 2–3; review objec-tive, 2; vision vs. reality, 2; WB policy paper,2; WB portfolio, 1

World Bank Institute (WBI), 174

Zimbabwe, 67Zoning, 19