private solutions for infrastructure in rwanda - isbn ...documents.worldbank.org/curated/en/... ·...

A C O U N T R Y F R A M E W O R K R E P O R T

THE WORLD BANK

PUBLIC-PRIVATE INFRASTRUCTUREADVISORY FACILITY

34417

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Private Solutionsfor Infrastructure

in Rwanda

Private Solutionsfor Infrastructure

in Rwanda

A Country Framework Report

The Public-Private Infrastructure Advisory Facilityand the World Bank Group

© 2005The International Bank for Reconstruction

and Development/THE WORLD BANK

1818 H Street, NWWashington, DC 20433

Telephone: 202-473-1000Internet: www.worldbank.org

E-mail: [email protected]

All rights reserved.

Manufactured in theUnited States of America.

1 2 3 4 07 06 05 04

The findings, interpretations, and conclusions expressed herein are those ofthe author(s) and do not necessarily reflect the views of the Board of Execu-tive Directors of the World Bank or the governments they represent.

The World Bank does not guarantee the accuracy of the data included inthis work. The boundaries, colors, denominations, and other informationshown on any map in this work do not imply any judgment on the part ofthe World Bank concerning the legal status of any territory or the endorse-ment or acceptance of such boundaries.

Rights and Permissions

The material in this work is copyrighted. Copying and/or transmitting por-tions or all of this work without permission may be a violation of applicablelaw. The World Bank encourages dissemination of its work and will normallygrant permission promptly.

For permission to photocopy or reprint any part of this work, please senda request with complete information to the Copyright 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 the Office of the Publisher, World Bank, 1818 HStreet, NW, Washington, DC 20433, USA, fax 202-522-2422, [email protected]

ISBN 0-8213-5965-7

Library of Congress Cataloging-in-Publication Data

Private solutions for infrastructure in Rwanda.p. cm.

“Published jointly by the Public-Private Infrastructure Advisory Facility and theWorld Bank”—Preface.

Includes bibliographical references.ISBN 0-8213-5965-71. Infrastructure (Economics)—Rwanda. 2. Public-private sector cooperation—

Rwanda. I. Public-Private Infrastructure Advisory Facility. II.World Bank.

HC875.Z9C374 2004363.6'0967571—dc22 2004053502

v

Preface ixAbbreviations and Acronyms xi

Executive Summary 1The Context for Reform 1Infrastructure Reform in Rwanda 1The Country Framework Report 2Summary of the Sector Reviews 2Cross-Cutting Issues 6Prioritized Recommendations 10

1. Introduction: Policies, Objectives, andthe Role of the Private Sector 11Background 11Objectives and Scope of the Country Framework Report 12Principles of Private Sector Participation 13

2. The Transport Sector 16Roads 16Airports 26Rail 28Water Transport 29Notes 31

3. The Energy Sector 32Sector Structure, Roles, and Responsibilities 32Policymaking, Planning, and Regulation 32Electricity 34The Gas Sector 35Investment Needs and Priorities 37Potential Sources of Investment Financing 38Opportunities for Private Sector Participation 38Issues 39Recommendations 40Notes 41

Contents

4. The Water and Sanitation Sector 42Sector Structure, Roles, and Responsibilities 42Policymaking, Planning, and Regulation 42Sector Performance 45Opportunities for Private Sector Participation 50Issues 53Recommendations 55Note 56

5. Common Issues in the Electricity and Water Sectors 57Management and Human Resources 57Commercial and Financial Management 57Management Contract 57Recommendations 58

6. The Telecommunications Sector 59Sector Structure, Roles, and Responsibilities 59Policymaking, Planning, and Regulation 60Sector Performance 64Opportunities for Private Sector Participation 66Issues 67Recommendations 69Note 69

7. Cross-Cutting Issues 70Institutional Capacity 70Local Financial Markets 72Regulation 72Accounting System 74Taxation System 74Legal System 75Bid Evaluation Process 79Project Selection, Prioritization, and Planning 79Government Policy Communication with the General Public 80Environmental Issues 80Other Cross-Cutting Issues 82Proposed Implementation Schedule for the Recommendations 82

Appendix AThe Environment 89

Existing Environmental Institutions, Policies, and Legislation 89Environmental Issues in the Transport Sector 92Environmental Issues in the Energy Sector 92Environmental Issues in the Water Supply and Sanitation Sector 95Environmental Issues in the Telecommunications Sector 96Ecotourism in Rwanda 97

Appendix BThe Legal System 98

Rwandan Land Law 98Customary Law 98Land Law Reform 98Rights of Ownership 99

Appendix CRoad Maintenance, Rehabilitation, and Reconstruction Costs 100

Contents

vi

Contents

vii

References 103

BoxesA.1 Requirements under European Commission Directive

97/11/EC for the Completion of Environmental ImpactAssessments for Transport-Related Infrastructure Projects 92

A.2 Requirements under European Commission Directive97/11/EC for the Completion of Environmental ImpactAssessments with Respect to Infrastructure Projects inthe Energy Sector 93

A.3 Requirements under European Commission Directive97/11/EC for the Completion of Environmental ImpactAssessments with Respect to Infrastructure Projects inthe Water and Sanitation Sectors 95

Figures2.1 Roads: Roles and Responsibilities 172.2 Airports: Roles and Responsibilities 263.1 Energy: Existing and Future Key Roles and Responsibilities 333.2 Benchmarking Data in Electricity 364.1 Urban Water Supply: Roles and Responsibilities 434.2 Rural Water Supply: Roles and Responsibilities 444.3 Unaccounted-for Water in African Water Utilities 464.4 Number of Connections of African Water Utilities 464.5 Average Price of Water Supplied by African Water Utilities 476.1 Telecommunications: Roles and Responsibilities 606.2 Telecommunications Networks in Rwanda 626.3 National Teledensity 646.4 Fixed-Line Waiting Lists 656.5 Fixed-Line Tariffs 656.6 Teledensity and Per Capita Gross Domestic Product 66A.1 The Vicious and Virtuous Cycles Relating Deforestation

to the Quality of Water 94

TablesES.1 Sector-Specific Opportunities 7ES.2 Cross-Cutting Constraints 81.1 Foreign Direct Investment in Rwanda 142.1 Extent of the Classified Road Network 182.2 Condition of the Classified Road Network 192.3 Air Cargo Rates for General Cargo 272.4 Commercial Traffic 302.5 Annual Traffic from Gisenyi to Cyangugu, 1985 302.6 Estimated Traffic, Gisenyi to Cyangugu, 1990 303.1 Electricity Provision in Rwanda 343.2 Rehabilitation of Assets in the Energy Sector 373.3 Proposals for New Generation Investment 373.4 Sources of Investment Financing 384.1 Electrogaz Water Tariffs 477.1 Implementation Schedule for Action Plan 83C.1 Expenditure and Investments Associated with Road Building,

Rehabilitation, and Maintenance of Various Types of Roadsin Rwanda 100

C.2 Expenditure and Investment Required to Rehabilitatethe Paved Road Network in Rwanda 101

C.3 Expenditure and Investment Required to Rehabilitate theUnpaved Main Road Network in Rwanda 101

C.4 Expenditure and Investment Required to Rehabilitate theUnpaved Secondary Road Network in Rwanda 101

C.5 Investment Projects: Base Scenario 102C.6 Investment Projects: Alternative Scenario 102C.7 Annual Road Maintenance Projects: Cost Summary 102

Contents

viii

ix

This Country Framework Report for Rwanda is oneof the first in a series of country reviews aimed at im-proving the environment for private sector involve-ment in infrastructure. Prepared at the request of theRwandan government, the report has three main ob-jectives:• To describe and assess the current status and per-

formance of key infrastructure sectors• To describe and assess the policy, regulatory, and in-

stitutional environment for involving the privatesector in those sectors

• To assist, through the above processes, policymakersin framing future reform and development strate-gies and potential private sector investors in assess-ing investment opportunities.This report was begun under the auspices of the

World Bank Group’s Infrastructure Action Program,with funding from the World Bank and the Japanesegovernment. It is being published jointly by the WorldBank and the Public-Private Infrastructure AdvisoryFacility, the multidonor technical assistance facility es-tablished in July 1999, which is carrying forward theprogram of Country Framework Reports begun underthe Infrastructure Advisory Facility.

The report was prepared by a core team led byLucy Fye and comprising Amadou Dem, GuidoRurangwa, Ibrahima Diong, Nikolay Mandinga, FatihaAmar, Marie Jeanne Uwanyarwaya, Marie-ChantalUwanyiligira, Serah Njoroge, and Aminetou Tidiani.The report also draws on inputs from various staff from

the World Bank and other development agencies, aswell as discussions with representatives of the privatesector.

The Country Framework Report process was sup-ported by a working group comprising representativesof the government, private sector, and bilateral donoragencies. Members of the working group from theRwandan government include Jean DemaceneNtawukuriryayo, Minister of Infrastructure; SamNkusi, State Minister of Energy and Communications;Munyanganizi Bikoro, State Minister of Water;Antoine Munyakazi-Juru, Ministry of Finance andEconomic Planning; Joseph Akilimali and MarleneNyirubutama, Privatization Secretariat; FrançoisXavier Havugimana, Makuza Kanamugire, AbrahamMakuza, Bruno Mwanzafunzi, J. Baptiste Ngwijabanzi,Straton Nzeyimana, Felicien Rukiriza, and SilasRuligana, Ministry of Infrastructure;Eugène Kivunan-goma, Multisector Regulatory Agency; Jean Kanya-muhanda, Central Public Investments and ExternalFinance Bureau; Patrick Rugumire, Ministry of Trans-port; Theogene Kayumba, Airports Authority; andAlfred Byigero, Gas Unit.

Members of the working group from the privatesector include Bart Gasana and Jules Ndenga, RwandaPrivate Sector Federation; Calvin Mitali, Rwanda In-vestment Promotion Agency; Jean Haguma, RwandaBar Association; Stuart King, John Mobsby, JonathanPell, David Phillips, Peter Rutamara, David Storer, andMatt Uzzell,Adam Smith Institute.

Preface

Representatives to the working group from bilat-eral donor agencies include Sandra Diesel, SwedishInternational Development Agency; Mark James,Department for International Development (UnitedKingdom); and Andy Karas, United States Agency forInternational Development.

This volume was produced by the World Bank’sOffice of the Publisher, which coordinated book de-sign, editing, and printing. Nora Ridolfi, Janet Sasser,and Thaisa Tiglao were instrumental in guiding bookproduction.

Preface

x

xi

AIM Alternative Investment MarketAIPA Africa Institute for Policy Analysis

and Economic Integration(South Africa)

AfDB African Development BankASI Adam Smith InstituteBADEA Arab Bank for Economic Development

in AfricaBOT Build, operate, and transferBUNEP Bureau National d’Etudes de ProjetCEPEX Central Public Investments and

External Finance BureauCFR Country Framework ReportDBFO Design, build, finance, and operateDEN Department of Energy, Ministry of

InfrastructureDFID Department for International

Development (United Kingdom)DRC Democratic Republic of the CongoDRFO Design, rehabilitate, finance, and operateECA Economic Commission for Africa,

United NationsEIA Environmental impact assessmentELG ElectrogazEMPA Environmental Management and

Protection ActGDP Gross domestic productGNP Gross national productICT Information and communication

technologyIFI International financial institution

IPP Independent power project KIST Kigali Institute of Science and

TechnologyMIED Ministry of Infrastructure’s Energy

DivisionMOI Ministry of InfrastructureMSR Multisector Regulatory AgencyMTN Mobile telephone networksMW MegawattNGO Nongovernmental organizationNUR National University of RwandaOBA Output-based aidOPEC Organization of Petroleum

Exporting CountriesPCM Pulse code modulationPDH Plesiochronous digital hierarchyPIP Public Investment ProgramPPI Private participation in

infrastructurePPIAF Public-Private Infrastructure

Advisory FacilityPRSP Poverty Reduction Strategy PolicyPSF Private Sector FederationPSP Private sector participationRFP Request for proposalsRIPA Rwanda Investment Promotion

AgencyRITA Rwanda Information Technology

AuthorityRMF Roads Maintenance FundRPSF Rwanda Private Sector Federation

Abbreviations and Acronyms

SEA Strategic environmental assessmentSIDA Swedish International Development

AgencySME Small and medium enterpriseUPEG Unité de Promotion et d’Exploitation

du Gaz du Lac Kivu (Department ofPromotion and Exploitation ofMethane Gas from Lake Kivu)

USAID United States Agency for InternationalDevelopment

VOC Vehicle operating costVSAT Very small aperture terminalWLL Wireless local loop

Abbreviations and Acronyms

xii

1

The Context for Reform

The systematic campaign of genocide and the civil warthat took place in April to July 1994 inflicted incalcu-lable harm on the social and economic fabric ofRwanda.One harmful aspect of the events of 1994 andtheir aftermath has been the enormous damage causedto the country’s infrastructure endowment.This dam-age reflects not only the direct effects of war and socialdisruption but also the indirect consequences in termsof (a) the loss of staff members in utility companiesand ministries—including many in senior managerialpositions—and the consequent loss of institutionalmemory, and (b) the destruction of records and the ef-fect this loss had on billing and payment systems andon financial performance.The ability of Rwanda’s eco-nomic infrastructure to fulfill the country’s needs hasbeen further undermined by the lack of regular main-tenance since 1994 and by the increased demandsplaced on the infrastructure as a result of the popula-tion resettlement, which has formed a vital part of therecovery process.

The achievements of the government and people ofRwanda in restoring the country have been remark-able. Political stability and security have been reestab-lished, and sound institutional and administrativeframeworks have been put in place. Despite difficultcircumstances, significant economic growth has beenachieved, and the economy is recovering to prewarlevels, although progress has slowed somewhat over thepast two or three years.

While much has been achieved, the government ofRwanda fully recognizes that much also remains to bedone. Its priority is not only to address the particularconsequences of genocide and war but also to tacklelonger-term structural difficulties already evident inthe prewar economy, particularly those caused by thepressures of population growth in a relatively smallcountry with an economy primarily dependent onagrarian activity.

Infrastructure Reform in Rwanda

The government’s vision for the development of thecountry is set out in its development agenda,Vision2020,which elaborates the economic goals that the gov-ernment aims to achieve and a broad strategy for theirrealization. One theme that lies at the heart of Vision2020 is the refurbishment and development of thecountry’s core economic infrastructure. This commit-ment to infrastructure improvement reflects the govern-ment’s recognition of the central role that reliable andaccessible infrastructure plays in supporting and enablingpoverty eradication and economic growth.

The progress that Rwanda has made since 1994 hasbeen achieved primarily through effective partnershipbetween the government and the international donorcommunity. Looking ahead, however, the governmentfully recognizes the increasing importance of theprivate sector’s contribution to realizing the country’sdevelopment agenda. A core theme of the govern-ment’s strategy for implementing Vision 2020 is the

Executive Summary

promotion and encouragement of private sector par-ticipation (PSP) and of private investment, includinginvestment by both foreign and local investors, in allsectors of the economy. Encouraging greater privateparticipation in infrastructure is accordingly a majorelement in the government’s policy for infrastructurerefurbishment, operation, and improvement.

In part, the importance given to private participa-tion in infrastructure reflects the need to harness finan-cial resources that substantially exceed what govern-ment and donors can be expected—or will be able—toprovide. At least as important, however, are the widerbenefits that flow from greater private sector involve-ment in the provision of infrastructure services. Signif-icant among these benefits are• The greater stimulus that private participation

provides—within appropriate market and regula-tory structures—to enhance efficiency in the use ofresources and avoid wasted cost and effort

• The potential that exists to promote the involve-ment of the Rwandan private sector and local com-munities in infrastructure, providing both liveli-hood opportunities to individuals and scope for thedevelopment of the indigenous business sectorand for the progressive enhancement of its wealth-generating capacity

• The greater potential that PSP offers for demand-led provision that is both more innovative and moreresponsive to economic and social needs than cen-trally planned development.Reforming the infrastructure sectors through PSP

is not simply a matter of selling a stake in public sectorassets to the private sector or otherwise bringing theprivate sector into publicly owned businesses throughconcessions or management contracts, although theseactions have an extremely important part to play. Infra-structure reform is also concerned with identifying andimplementing structural changes to create competitivemarkets; implementing institutional reforms to providelegal and regulatory frameworks that ensure fair andeven-handed protection of the interests of users,providers, and the public; and assessing the appropriaterole for the public sector.

Achieving success in infrastructure reform alsoinvolves much more than identifying appropriate sec-tor policies and the opportunities that exist to promotePSP within these policy frameworks. It is also, crucially,

about effective implementation, which requires clearlyappraising priorities and focusing scarce resources ontheir achievement. It also requires careful planning,allocation (and full acceptance) of institutional respon-sibilities, establishment of an appropriate enablingenvironment, effective monitoring and review ofprogress, and genuine political will. Moreover, it relieson removing or overcoming broader obstacles thatcut across more than one sector and whose effectmay, indeed, extend more widely through Rwanda’seconomy.

The Country Framework Report

The Rwanda Country Framework Report (CFR) sets outa review of Rwanda’s infrastructure in the transport,energy, water and sanitation, and telecommunicationssectors. Although the span of the CFR is broad, it isfocused in particular on the potential that exists toinvolve the private sector in infrastructure and thepolicies and actions that are necessary to bring thisinvolvement about.

The CFR is an objective assessment of the condi-tion of Rwanda’s infrastructure sectors and of theinstitutional and policy frameworks associated withthem.Although the government has achieved much, itrecognizes fully the room that exists for improvement.The CFR is not a sales prospectus for Rwanda’s infra-structure. It is, however, intended to provide a clearroute map for infrastructure sector reform and to high-light both the opportunities that exist for the privatesector and the role that the donor community can playin helping the government realize its priorities ininfrastructure.

The remainder of this executive summary high-lights the main themes that have emerged from theCFR exercise for each infrastructure sector and thensummarizes the core cross-cutting issues that have beenidentified. It concludes by outlining the main CFRrecommendations.

Summary of the Sector Reviews

TransportAs a landlocked country with mountainous terrain,Rwanda faces unusually difficult problems in relationto both national and international transportation.

Private Solutions for Infrastructure in Rwanda

2

Inadequacies in its transport infrastructure impose sig-nificant additional costs on the country’s economy andrepresent a serious impediment to improving percapita incomes. Because of such inadequacies, thepoorest rural communities face considerable problemsgaining access to markets for their produce. Inferiortransport infrastructure—particularly, the inadequaterural roads—also affects the ability of the poor to gainaccess to key services, such as health and education,and represents a serious obstacle to alleviatingpoverty. Moreover, deficiencies in Rwanda’s transportsystems spill over into the other infrastructure sectors,adding to maintenance costs and reducing the qualityof service.

Roads Overall responsibility for road policy rests withthe Directorate of Roads, which is part of the Ministryof Infrastructure.Under the government’s decentraliza-tion policy, responsibility for maintaining the unpavednational road and communal road network is, however,being transferred from the central government to theprovincial governments. Maintenance of paved roads isnow carried out by the private sector. The govern-ment’s policy for the road subsector has not yet beenfully articulated, but its core goals are clear.Their cen-tral thrust is (a) to enhance Rwanda’s integration intothe regional economy by improving the condition—and subsequently the extent—of the country’s nationalroad system, and (b) to improve the availability andquality of local road infrastructure in order to promoterural development by providing better access to mar-kets. The promotion of PSP in road rehabilitation,maintenance, and development is a central objective inpursuit of these goals.

The condition of the whole road network hasdeteriorated substantially in the past decade, resultingin a significant cost to the economy. As far as can beascertained, the condition of bridges, viaducts, and cul-verts mirrors that of the roads of which they form apart.Vehicle operating costs in Rwanda are high, as adirect consequence of the poor condition of the roads,and access to both product markets and essential pub-lic services is impaired, especially in poor rural areas.

There are significant opportunities for PSP in theroad rehabilitation,maintenance, and development pro-gram. Initially, a significant proportion of this activitywill be passed to the private sector through 5- or

10-year concession contracts. Attracting foreign con-tractors is likely to be a first priority, but after takingsteps to help build local capacity, the program shouldalso provide a real opportunity for the involvement ofRwandan businesses.

Although significant PSP opportunities exist,important issues must be addressed if these opportuni-ties are to come to fruition.The Roads MaintenanceFund (RMF) is at present underfunded, there is a lackof clear prioritization of roads, and the institutionalcapacity to manage a contract program is wanting.Furthermore, a direct consequence of the currentdecentralization policy is likely to be that overallmaintenance of unpaved roads—and its quality—willbecome very difficult to control.

Airports Overall responsibility for the airport subsectorrests with the Ministry of Infrastructure. The state-owned Airports Authority carries out both planningand management.There are two airports and five air-fields in Rwanda, including one international airport,Kanombe Airport at Kigali. The government’s policyobjectives are to promote regional integration, to es-tablish Kanombe as a regional hub, and to promote thedevelopment of air transport.

Traffic at Kanombe is significantly below the capac-ity of the airport. Kanombe is currently served by onlyone European airline, four African airlines, and onelocal airline. As with road transport, the cost of airtransport is high owing to the lack of competition andthe small size of the market.

There seems little potential for significant PSP inthe airport subsector in the short term. A number ofoptions should be considered in the medium term,including contracting airport management out to theprivate sector or privatizing the Airports Authorityitself, if the economy develops and passenger numbersgrow. In the short term, it may also be possible tointroduce PSP options on a smaller scale—for exam-ple, contracting airport car parking out to the localprivate sector.

The question of whether there is a need to developa second runway at Kanombe has also been raised.While a second runway would clearly raise capacity atthe airport, the question of whether the associatedexpenditure would be justified depends on expectedfuture air traffic. Seeking to develop this project as a

Executive Summary

3

PSP scheme would be one way to test the market viewof the risk that future airport revenue might not ade-quately remunerate such an investment. A similarapproach might be adopted at Kibuye, where it hasbeen argued that an improved strip is needed to pro-mote tourist traffic.

Careful design would be necessary to ensure thatprivate investors would be remunerated on a basis thatreflects any future traffic growth that is achieved.Again, the willingness of the private sector to becomeinvolved in such a scheme would provide a clear test ofthe market’s view concerning achievable growth intourist traffic at Kibuye.

Rail Rwanda is currently not served by any railway sys-tem.The government’s key policy objective for rail isto investigate the feasibility of establishing a railway toconnect to the ports of Dar es Salaam or Mombasathrough one of the other regional railway systems,withthe aim of reducing costs for the transport of goods inbulk. The government is also keen to consider the pos-sibility of establishing a southern transport corridorlinking the countries of the Great Lakes regionwith South Africa, as part of the Great Lakes RailwayProject. The main concern, however, relates to trafficpotential. There is significant uncertainty aboutwhether what would be a large investment could bejustified given the current and projected volume ofRwanda’s exports and imports.

Water Transport Although the water transport subsectoris currently very limited in extent, there are potentialopportunities for PSP in the further development ofwater transport on Lake Kivu. A serious constraint isthe current absence of any boatyard facilities for con-struction or maintenance on the Rwandan shore ofthe lake.

EnergyEnergy sector policy is the responsibility of theDepartment of Energy, within the Ministry of Infra-structure. The provision of electricity (and water) inthe urban districts of Rwanda is the responsibility ofElectrogaz (ELG), a parastatal organization. Microgen-eration and small regional networks may be owned andoperated at the community level.The Department ofPromotion and Exploitation of Methane Gas from

Lake Kivu (Unité de Promotion et d’Exploitation duGaz du Lac Kivu, or UPEG) is responsible for devel-oping the methane resource at Lake Kivu, which—ifmanaged appropriately—represents a natural resourceof enormous potential value to Rwanda and its long-term development.

The government of Rwanda has not prepared andpublished a single, comprehensive, and coordinatedpolicy statement and program for the energy sector,although it has prepared a general strategy for thereform and development of the sector. The earlyadvancement of such a policy statement is of greatimportance in helping shape the future development ofthe sector and in providing a clear framework withinwhich PSP can flourish.The focus of the government’senergy policy is to promote activities that will increaseaccess to electricity and provide a quality and cost-effective service while simultaneously ensuring thefinancial viability of economic agents engaged in pro-viding energy services and protecting the environment.The government plans to implement a range of policymeasures to achieve this objective. These measuresinclude promoting competition and PSP in the sectorthrough regulatory change and the transfer of manage-ment responsibility for ELG to a private operator undera management contract.The government also plans topromote rural electrification, in terms of both networkextension into rural areas and local power generation.

Approximately 4 percent of the population withinELG’s urban area of operation is estimated to be con-nected to the network, an exceptionally low level ofcoverage. Rwanda’s electricity generation portfolio isdominated by one hydroelectric plant, making it vul-nerable to both climatic and environmental fluctua-tions. More than half of Rwanda’s electricity needs aremet through imports or through the purchase ofBurundi’s share of the output of the Sinelac operation(owned in equal shares by Burundi, the DemocraticRepublic of Congo, and Rwanda). Rwanda suffersfrom a deficit in generation capacity, which is exacer-bated by poor standards of maintenance. Deficienciesin the transmission and distribution systems combineto undermine the quality and reliability of supply.Bothtechnical and commercial system losses are high.

Operational assets in the gas subsector are currentlylimited to a single, small-scale, methane gas extractionfacility on Lake Kivu.


4

There are a number of major opportunities for PSPin the energy sector. If the ELG management contractis successful in meeting its objectives, it should be pos-sible to establish more extensive forms of PSP in urbanelectricity, which would involve a significant fundingcommitment by the private sector. Issuing a concessionfor a methane gas extraction plant at Lake Kivu and anassociated generation facility offers the prospect offilling the urgent need for additional power generationcapacity and represents another key opportunity. PSPwill certainly be required—in combination with exter-nal subsidies from the government or the donorcommunity—if a far-reaching rural electrificationprogram is going to be implemented.

Key to the success of these potential opportunitieswill be the development of a clear policy frameworkfor the energy sector setting out, among other things,the government’s objectives and the structural and oper-ational evolution that it intends and the competitivemodel that it will adopt (covering both electricity andgas). Also critical to success will be the introduction ofnew sector laws to provide an enabling framework forthe execution of this policy and the introduction ofeffective regulatory arrangements, rules, and mecha-nisms. The success of the ELG management contractand the effectiveness of its performance will be ofcentral importance too. It is vital that the function,responsibility, and jurisdiction of the Multisector Reg-ulatory Agency—as far as energy sector operation isconcerned—be specified as soon as possible. Great carewill need to be exercised in negotiating and executingan agreement to develop the methane gas resource forpower generation purposes: the terms of this agree-ment will affect the potential for competition and PSPin the energy sector for many years. A rural electrifica-tion plan—together with measures to promote small-scale, local private sector and community involvementin electricity supply—is also a key priority.

Water and SanitationOverall policy responsibility for water supply, sanita-tion, and water resource management resides with theMinistry of Infrastructure, in the Department of Waterand Sanitation. Operational responsibility for the watersupply belongs to ELG in urban areas and to the dis-tricts in rural areas. Responsibility for sanitation serv-ices resides at the district level in both urban and rural

areas, although in practice almost no service is actuallyprovided.

Overall access to potable water in Rwanda is poorby African standards.The great majority of the popu-lation in both rural and urban areas is served throughone form or another of communal facility; individualconnections are uncommon. Sanitation services arealmost entirely self-provided and are agreed to be of agenerally poor standard. As recognized both in theUnited Nations Millennium Declaration and at the2002 Johannesburg summit, improvement in the pro-vision of adequate water supply and sanitation serv-ices has a more direct effect on human health andwelfare than improvement in any other infrastructureservice. Although the direct effect of water and sani-tation services on economic growth may be morelimited than that of other infrastructure sectors, theycontribute directly to the well-being of the poorest insociety.

The government’s core policy objectives for thewater supply and sanitation sector are to improve theprovision of water, extend the water supply network,and increase access to sanitation services, using meansthat promote technically and financially viable projectsbased on strong community participation, as well as tostrengthen capacity at both the central governmentand the district levels.These policy objectives and thespecific means to achieve them are, however, not cur-rently set out in any formal policy statement, althougha water supply policy document is being prepared.

ELG’s technical and commercial performance hasbeen weak, with high levels of both physical and ad-ministrative losses, in part reflecting the considerableresource problems that the company faces and thelegacy of the events of 1994. A key aim of the ELGmanagement contract is to address those performanceweaknesses. The available evidence suggests that theprice that Rwandans pay for their water is unusuallyhigh by regional standards, despite the generally poorlevel of service that they receive.

The way in which Rwanda manages its waterresources has critical implications for water supply andfor sanitation and vice versa. The government recog-nizes that serious issues must be addressed. Kigali suf-fers from very difficult water resource problems thatcan in part be attributed to serious deterioration inthe quality of raw water resources. A clear policy

Executive Summary

5

framework to address water resource managementissues is in place but is not being implemented.

There are substantial opportunities for PSP in thewater supply sector.They include (a) establishing newwater sources to meet the current water supply deficitin Kigali, (b) deepening the extent of private involve-ment in Electrogaz (as discussed in relation to theenergy sector), (c) promoting the development of net-worked rural service through complementary low-costnetworks, (d) introducing arrangements to promotedemand-led private participation in rural water supply,and (e) enhancing the role of urban standpipe opera-tors, with the aim of improving their livelihoods andreducing urban water prices. A significant feature of anumber of these opportunities is the extensive scopethat they provide for—indeed, their reliance on—community participation and the involvement oflocally based small businesses and entrepreneurs.

TelecommunicationsTelecommunications development, which is the policyresponsibility of the Department of Communicationswithin the Ministry of Infrastructure, plays a centralrole in the government’s vision of economic develop-ment in Rwanda—a vision in which information andcommunication technology (ICT) takes the lead. Un-less there is rapid and sustained expansion of Rwanda’stelephone network, which is currently one of the leastextensive in Africa, it is clear that this vision will beimpossible to realize. Privatization and liberalizationof telecommunications have been shown—in nu-merous countries at a variety of stages of economicdevelopment—to represent the core requirement ingalvanizing rapid increases in access to modern tele-phony services and in making quick improvements tothe range and quality of services offered to consumers.

There is currently one parastatal fixed-line operator,Rwandatel, and one major mobile operator, Rwanda-cell, which is jointly owned by Rwandatel, MobileTelephone Networks (MTN), and a local investmentvehicle. The government has embarked on a strategyto open the sector fully to competition, to unwindRwandatel’s cross-holding in Rwandacell, and to sell amajority stake in Rwandatel,which is wholly owned bythe government, to a strategic investor. Some key ele-ments of the regulatory framework that will be needed ifthese policies are to be successful remain to be finalized.

Rwandatel’s performance has been poor across awide range of service attributes. The legacy of 1994had a particularly deleterious effect on the company’sperformance. Compared with other African telecom-munications service providers, Rwandatel’s marketpenetration is exceptionally low, with only 0.27 fixedlines installed per 100 people, although this deficit isbalanced somewhat by the extent of connection to themobile network.The overall teledensity of 1.1 in 2001is, nevertheless, low by regional standards.

With successful restructuring and liberalization,there will be significant opportunities for PSP inRwanda’s telecommunications sector, both throughinvestment in Rwandatel and through new entry tothe market. For these opportunities to materialize,however, it is critical that an effective and fair regula-tory framework be established as soon as practicablypossible, in particular covering interconnection andlicensing arrangements. Unless an open and fair com-petition regime is established quickly, Rwandatel canbe expected to continue to underperform—with orwithout PSP—and private operators and capital willnot be drawn into the sector through the entry of newcompetitors.

There are further opportunities for PSP in ICT,particularly in the context of Rwanda’s strategy ofICT-led development. Most of these opportunities arerelatively small in scale and particularly well suited toexploitation by the country’s indigenous businesses.

Cross-Cutting Issues

Some of the most important constraints on the devel-opment of Rwanda’s infrastructure—and, in particular,on the exploitation of PSP to improve infrastructureservices—are not specific to individual infrastructuresectors but act across a range of sectors.They may alsohave an important affect on private sector developmentin the wider economy. These cross-cutting issues aresummarized below:• Institutional capacity. Insufficient institutional capac-

ity in government and parastatal organizations is aconstraint that is commonly encountered in devel-oping countries. In Rwanda, however, capacityweaknesses in both the government and the para-statals have been significantly exacerbated by theimpact of the genocide. Within the government,


6

Executive Summary

7

Sector Opportunity Main project Supporting policies and actions

First-priority recommendations

Other priority recommendations

Table ES.1 Sector-Specific Opportunities

Proceed with Lake Kivu gasconcession agreement.

Move all rehabilitation andmaintenance contracts to a 10-year concession basis.

Conduct feasibility study of flexible, demand-led fundingapproach to rural electricity supply, with competitive bidding for capital subsidies from a ruralelectricity supply fund.

Update and reconsider 1985 study.

Conduct feasibility study of flexible,demand-led funding approach torural water supply, with competitivebidding for capital subsidies from arural water supply fund.

Conduct full feasibility study andfinancial appraisal, including reviewof appropriate PSP model.

Energy

Transport

Energy

Transport

Water andsanitation

Exploit potential of Lake Kivu gas tomeet electricity generation needs.

Attract private sector participation inroad rehabilitation and maintenance.

Promote partnerships in provision of rural electricity service to harnesscombined potential of governmentand donor funding with local privatesector, community, andnongovernmental organizations(NGOs).

Exploit potential for PSP indevelopment of water transportationon Lake Kivu.

Promote partnership arrangements inprovision of rural water services toharness combined potential ofgovernment and donor funding withlocal private sector, community, andNGOs.

Address Kigali water supply problemsthrough PSP scheme to developRuhengeri-Kigali long-distancepipeline.

Make project structure and agreements consistentwith long-term energy policy framework and newsector laws (for example, regarding future energysector reform).

Ensure required experience and expertise isengaged to conclude an agreement that will beattractive to banks and that will achieve full valuefor Rwanda.

Return control of maintenance to Directorate ofRoads of the Ministry of Infrastructure.

Prioritize roads in order of their importance to theeconomy.

Prepare a fully costed 10-year road rehabilitationand maintenance plan.

Enforce axle-load restrictions.

Consider penalties for vehicle-causedenvironmental damage.

Strengthen law to protect and enforce rights of way.

Immediately increase all Roads Maintenance Fundcharges and improve collection procedures.

Consider establishing Directorate of Roads as anindependent roads agency.

Agree with donors on a financing plan for roadrehabilitation plan.

Establish close cooperation with donors and withboth national and international NGOs.

Ensure that ELG is able to participate in theprovision of rural electricity services and able toobtain capital subsidy where efficient and effective.

Structure rural energy with measurable objectives,clear funding policy, and prioritized action plan.

Update 1986 boatyard study.

Establish close cooperation with donors and withboth national and international NGOs.

Ensure that ELG is able to participate in theprovision of rural water services and able toobtain capital subsidy where efficient and effective.

Carry out environmental impact assessment.


8

Category Constraint Main action Supporting policies and actions

First-priority issues and recommendations

Other priority issues and recommendations

Table ES.2 Cross-Cutting Constraints

Support provision of additionalmicrofinance facilities through local banks.

Encourage and support local banks inproviding medium- and long-term bankdebt and financing for company start-ups.

Amend land law.

Review company (and concession) law toensure private sector investor can protectits ability to manage its investment.

Amend 1998 statutory disputes resolutionmethods to accord with internationalpractices, including measures to improvetransparency.

Make all laws relevant to the privatizationprocess available in official translations intoEnglish.

Procure technical assistance for MSRdepartment heads with measurable targetsfor knowledge transfer.

Review composition and appointment ofRegulatory Board.

Amend the Regulatory Law to makeRegulatory Board role supervisory andtransfer executive authority to MSRmanaging director.

Evaluate and implement effectiveinstitutional and procedural arrangementsfor appeal against regulatory decisions.

Abandon confrontational approachcurrently adopted by tax authorities.

Align tax incentives with regional practice.

Introduce reforms of accounting standardsand procedures.

Government, with donor assistance, shouldconsider establishing a small informal equitytrading market.

Set up technical assistance and trainingprograms to train bank staff in medium-and long-term lending, project financetechniques, and provision of financing forstart-up companies.

Local financialmarkets

Legalframework

Regulatoryframework

Tax system

Accountingsystem

Local financialmarkets

Availability of microfinance facilities

Lack of a medium- and long-termdebt market

Insufficient security of tenure overland acquired by investors for thepurposes of PSP schemes

Conflict between certain provisionsof company legislation andprivatization requirements

General provisions for disputeresolution that are unlikely to beacceptable for outside investors

Laws that are not fully accessible tointernational investors

Risks that demands on MultisectorRegulatory Agency (MSR) will exceedcapacity in the near term.

Lack of MSR institutionalindependence from government

MSR internal governancearrangements that impede efficient,timely decisionmaking

Inadequate arrangements for appealagainst regulator decisions

Overaggressive taxation procedures

Weak incentives, by internationalstandards

No national accounting standardssystem and lack of confidence oflending institutions in corporateaccounting information

Lack of opportunities for raisingprivate equity

Inadequate local bank expertise,knowledge, and experience in projectfinance techniques

Establish predominance of independentmembers.

Adopt open and transparent publicselection procedures (for example, throughpublic selection committee).

Consider a tax holiday for investors of fiveyears from start of operations.

Increase carryforward period for tax lossesto five years.

Establish international accounting standards.

Allow only qualified accountants to auditaccounts.

Executive Summary

9

Category Constraint Main action Supporting policies and actions

Other issues and recommendations

Table ES.2 (Continued)

Consider upgrading existing laws onconcession contracts.

Amend Privatization Law.

Revise importation procedures.

Clarify government environmental policiesand update National Environmental ActionPlan.

Update and improve framework legislation,and eliminate duplication with other draftlaws.

Use environmental impact assessments askey tools to identify and mitigate impactsarising from specific forms of infrastructuredevelopment.

Introduce range of reforms to improvetreatment and handling of liquid and solidwastes.

Review current system and speed upprocedures to remove unnecessary delaysand reduce project costs.

Review current procedure to ascertainwhether additional checks are necessary,ensure that procedure is not misused, andsee that the right balance is maintainedbetween needs and resources.

Introduce reforms to improve governmentcommunications.

Legalframework

Tax system

Environmentalissues

Bid evaluationprocess

Projectselection,prioritization,and planning

Policycommunication

Absence of laws dealing specificallywith concession contracts; build,operate, and transfer arrangements;and so forth.

Unduly restrictive privatization lawwith respect to PSP schemes notinvolving sale of assets

Cumbersome and slow importationprocedures

Environmental policy, law, and practicethat fall short of best practices fordeveloping countries

Slow and unwieldy procedures forevaluating bids

Possibility of misuse of current system

Weaknesses in communication withgeneral public and with parastatals

Consider single-agency responsibility forpublicizing policy for private participation ininfrastructure.

Improve communication between ministriesand parastatals through allocation andclarification of responsibilities.

capacity constraints are also magnified in the imme-diate term by the effects of the country’s decentral-ization policies.

• Local financial markets. Constraints arise in relationto, among other things, the lack of local markets toraise both equity and medium- and long-term bankdebt, the lack of sufficient local microfinance facili-ties, and the lack of banking staff capability in rela-tion to project financing.

• Regulation. Cross-cutting issues arise in connectionwith the urgent short-term need to build regulatory

capacity, the lack of independence of the Multisec-tor Regulatory Agency, and the inadequate arrange-ments for appeals against the regulator’s decisions.

• Accounting system. Cross-cutting constraints onRwanda’s accounting system include the lack of anational accounting standards system, the lack ofconfidence among lending institutions in corporateaccounting information provided to support loanrequests, and problems with tax payments arisingfrom Revenue Authority disputes of corporate taxand profit calculations.

• Taxation system.Cross-cutting issues include overag-gressive taxation procedures and weak tax incen-tives, by international standards.

• Legal system. Issues include insufficient security oftenure for investors over land acquired in PSPschemes; conflicts between provisions in the com-pany legislation and privatization requirements;need for amendment of dispute resolution provi-sions; weak sector law in the telecommunications,water and sanitation, and road sectors; and inacces-sibility of key laws to international investors.

• Bid evaluation process. Procedures for evaluating bidsare slow and unwieldy.

• Project selection, prioritization, and planning. Thedesign of the current system may allow it to bemisused.

• Government policy communication with the general pub-lic. Weaknesses exist in communication with thegeneral public and with parastatals.

• Environmental issues. Environmental policy, law, andpractice fall short of best practices for developingcountries.

Prioritized Recommendations

Table ES.1 summarizes, by priority, the main sector-specific PSP opportunities identified in the course ofthe CFR exercise. It identifies projects and supportingpolicies or actions required in order to exploit thoseopportunities. Table ES.2 sets out cross-cutting con-straints that have been noted and, again, identifies therange of actions necessary to address these constraints.


10

11

This Country Framework Report (CFR), prepared bythe government of Rwanda with the support of theWorld Bank and the Public-Private InfrastructureAdvisory Facility (PPIAF), is an infrastructure policy as-sessment and a review of the potential for private sectorparticipation (PSP) in providing public infrastructureservices. It presents a balanced assessment of the currentstate of the country’s infrastructure, policies, and policyreform, together with practical recommendations toimplement and achieve government objectives for in-frastructure rehabilitation and development with theparticipation of the private sector. The CFR is in-tended to provide a key roadmap to assist both publicand potential private sector participants in Rwanda’s in-frastructure development.

Background

The transition from conflict and emergency to sus-tainable development and growth is now under way.The economy is recovering to the 1990 prewar level,although there has been a slowdown over the pasttwo or three years. The government has adopted anumber of long-term perspectives for the develop-ment of the country, which are set out in its develop-ment agenda, Vision 2020. This vision centers oncritical policy strategies that are designed to generatehigh levels of growth and rapid poverty reduction.Vision 2020 is the government’s mission statementand ultimate national goal. It represents a broad, long-term view of the desired economic position of the

country to be reached by 2020.These are the six keycriteria for Vision 2020:• Strengthening and maintaining good governance• Transforming the agricultural sector into a high-

value and high-productivity sector• Developing human resources• Developing a knowledge-based service sector with

emphasis on information, communication, andtechnology

• Reducing risks and costs of doing business by estab-lishing an enabling environment and developing therequisite infrastructure

• Promoting the entrepreneurial class and regionalintegration.The government realizes that it does not on its

own—even with the considerable assistance of thedonor community—have the resources to implementVision 2020. It has, therefore, embarked on a strategy toencourage the involvement of the private sector on apublic-private partnership basis, to implement Vision2020. Measures have already been taken. Structural re-forms have focused on opening up the economy andenhancing the economic and regulatory environmentfor private sector activity. To promote private sectordevelopment, the government of Rwanda has simpli-fied business licensing requirements, revised the laborcode, and established the Rwanda Investment Promo-tion Agency to facilitate investment and business devel-opment. In 2000, the government-controlled Chamberof Commerce was replaced by an independent PrivateSector Federation, which represents private sector

1 Introduction: Policies, Objectives, andthe Role of the Private Sector

interests in its dialogue with the government.Much stillremains to be done, but the government has made agood start.

The status of Rwanda’s infrastructure is the conse-quence of the civil war and the genocide in 1994,which destroyed major economic sectors, particularlythe infrastructure sectors. With the support of thedonor community, some infrastructure has been reha-bilitated during the past 10 years, but much remains tobe repaired. In addition, the infrastructure has not beensufficiently maintained, and the quality of the country’sinfrastructure has declined in the past few years. Poorinfrastructure constitutes a major barrier to the coun-try’s economic activity, reduces competitiveness of ex-ports, and discourages investment.

The country’s poor infrastructure also significantlyimpedes the government’s central policy objectives ofreducing poverty and improving social welfare.This isespecially true in the rural community. As stated above,the government does not have sufficient resources toaddress the inadequacies of its national infrastructure.The government has therefore recognized thatRwanda, like many other developing countries, mustharness the resources and skills of private sector infra-structure service providers and operators if it is tosuccessfully meet the challenge of transforming itsinfrastructure and realizing the goals of Vision 2020.

Successfully involving the private sector in the con-struction, maintenance, expansion, life-cycle replace-ment, and improvement of infrastructure confrontsdeveloping countries like Rwanda with the need toachieve fundamental change in the financial, regula-tory, and operating environments within which theinfrastructure sector functions. Because of a longstand-ing tradition of public sector provision, operation, andfinancing of core infrastructure services, the institu-tional and legal frameworks of developing countrieslike Rwanda are generally not well adapted to theneeds and priorities of the private sector. Furthermore,administrative,managerial, and resource capacities con-strain the ability of the public sector to handle thechange to PSP in providing infrastructure services.

With these facts in mind, the government ap-proached the World Bank and PPIAF for assistance.The PPIAF recommended the preparation of a CFRon Rwanda’s infrastructure to provide a roadmap ofhow the government and the people of Rwanda can,

together with the private sector, best rehabilitate,main-tain, and develop the country’s infrastructure.

In addition to this CFR, the World Bank is alsocurrently assisting Rwanda with three infrastructureprojects: the Energy Sector Rehabilitation and UrbanWaste Management Project, the Transport Sector Pol-icy, and the Rural Water Supply and Sanitation Project.

Objectives and Scope of the CountryFramework Report

A central purpose of this CFR is to set out guidelinesfor implementing the government’s objectives andpolicies.The CFR will specifically look at the follow-ing areas:• Identify where infrastructure development or im-

provement can contribute most effectively to further-ing the wider government economic policy objectives,including poverty reduction.

• Ascertain where and how the private sector canmost fruitfully be involved in providing infrastruc-ture services.

• Develop practical policies to attract the privatesector to participate in providing infrastructureservices.

• Identify significant obstacles to attracting and retain-ing PSP in providing infrastructure services.

• Review the provision of infrastructure services to thepoorer members of society and find ways of improv-ing these services.

• Review the involvement of small- and medium-size enterprises in the delivery of infrastructureservices and investigate ways of increasing thisinvolvement.

• Target the lending support of the World Bank anddonor agencies to meet agreed objectives in theshort-, medium-, and long-term improvement anddevelopment of infrastructure, which are essentialto the economic development and poverty reduc-tion in Rwanda.

The CFR is organized as follows.The present chaptersets out the government’s overall objectives—the keypolicy content for this report—and the general infra-structure sector policies that fit within the overallobjectives. It also provides information about the invest-ment climate in Rwanda. Chapters 2 through 6 analyzeeach infrastructure sector, highlighting each sector’s


12

current performance and key issues.Chapter 7 analyzescross-cutting issues that arise across all infrastructuresectors and that can have wide-ranging effects on PSP.The appendixes provide additional information forclarifying and detailing matters, such as contacts madeduring the course of the CFR exercise and the mem-bership of the working group. Specific appendixes arealso devoted to more detailed discussions of environ-mental and legal issues.

Principles of Private Sector Participation

The government, supported by the World Bank andPPIAF, has adopted a policy to encourage the privatesector, both local and foreign, to participate in the re-habilitation, maintenance, and construction of thecountry’s infrastructure, as well as in the provision ofpublic infrastructure services.The targeted sectors aretransportation, energy, water and sanitation, and tele-communications. In addition, the government seeksthe support of the donor community in its policy ofrestoring the country’s infrastructure to an acceptablelevel on a coordinated basis instead of through the un-coordinated programs of the past.

To implement such a policy requires an under-standing of what motivates the private sector and pri-vate investors in infrastructure projects. In relation tointernational investment, it is important to consider thepolicies and institutional arrangements needed to dif-ferentiate Rwanda from other developing countriesfacing similar problems. First, in seeking to promoteinternational PSP in infrastructure,Rwanda is compet-ing in what is effectively a worldwide market. Second,the private sector and private investors require a stablepolitical and economic environment.Third, the privatesector and its investors will seek to earn a profit marginor investment return commensurate with the risks theyconsider they will be taking.

Consideration of the potential for PSP in Rwanda’sinfrastructure should not, however, be restricted to thescope of international investment in larger-scale proj-ects. Especially in relation to rural infrastructure and tothe improvement of core services (and poverty reduc-tion) for the 80 percent or more of Rwandans wholive in rural communities, the development of policiesand programs to promote small-scale indigenous PSPis likely to be of equal importance.

International InvestmentA worldwide market and a stable environment meanthat the private sector will only participate if it canearn an acceptable return for the risks it is taking; if itcan manage the work with a lack of political interfer-ence, provided that it obeys the laws of the country inwhich it is operating; and if the country offers attrac-tive opportunities to do business and invest. Thecountry must have an acceptable commercial law envi-ronment, an acceptable tax system, and an acceptablesystem for foreign service providers and investors totransfer their profits and dividends and any net disposalof their investment to their home countries within areasonable time frame. Profit margins will vary fromproject to project, depending on a number of factors,and will need to be negotiated as part of the pricing forany particular project. Regarding investment returns, arough guide is that investors in developing countriesseek returns in their own countries between 12 to14 percent per year before taxes.Although during thecurrent worldwide recession not many investors areachieving those returns, nevertheless they remain thetarget. For a project in a developing country, the re-quired return would increase to a minimum of 18 to20 percent per year before taxes, assuming that theoverall tax rate is no more than that normally payablein their own country, which is usually between 30 and40 percent. Such a return would normally require adouble tax agreement to be in place between Rwandaand the home country of the service provider orinvestor. In all other circumstances, the investment re-turn would have to be higher to mitigate the differenttax rates.

International private sector investors and operatorsalso usually want to insure against political risk.Thereare several options for doing so.One is to use the WorldBank’s political risk guarantee facility, available throughits Project Finance and Guarantee Group and sup-ported by the government of Rwanda. Another optionfor investors is to use the facilities offered by the Mul-tilateral Investment Guarantee Agency, which is part ofthe World Bank Group. Some national export creditagencies also offer investment insurance includingpolitical risk to resident corporations. Finally, there isinvestment insurance available by way of the interna-tional insurance market through centers such as Lloydsof London, although this insurance is often short term

Introduction: Policies, Objectives, and the Role of the Private Sector

13

only. The World Bank has recently assisted in estab-lishing the African Trade Insurance Agency, which isbased in Nairobi, to provide insurance services to aidthe development of regional trade.

Foreign direct investment figures for Rwanda since1979, shown in table 1.1, are useful as a benchmark toassess future progress. Rwanda has yet to achieve pre-war levels of investment.

Local InvestmentLocal private sector investors that benefit from moreintimate knowledge of local markets and customs arebetter able to protect themselves from some of the risksfaced by the international community.These investorswill often be prepared to become involved in PSPschemes that international investors would not con-template. Local businesses, however, are also limited intheir ability to participate in such schemes because oftheir lack of technical expertise, particularly in relationto the completion of larger and more complex proj-ects, and because of their lack of access to financing,even at the microlevel necessary to implementextremely small-scale and localized projects.

Aid and Pro-Poor Private Sector ParticipationIf PSP is to be genuinely pro-poor, it may be necessaryto provide subsidies to ensure that the financial burdenof serving the needs of the least well-off customers—and the risks that are associated with such groups—donot deter private sector involvement. In some cases, itmay be possible to ensure that the poorest citizens areable to afford a basic level of service by applying tariffstructures that incorporate a significant element ofcross-subsidy from high-volume users of a particularservice to those who use it in relatively small quanti-

ties. There are well-recognized drawbacks with suchan approach, however. Disproportionately high chargesto large users—particularly to business and industrialcustomers—may deter them from, for instance, usingwater in productive ways that would otherwise havepromoted economic development. Moreover, a tariffsystem that forces an operator to serve the least well-offon a loss-making basis while adding to the profitabilityof serving the better-off will inevitably create incen-tives for the operator to focus its attention on the richwhile ignoring the poor. Despite these potentialdrawbacks, some element of cross-subsidy in tariffs isgenerally acceptable as long as it is kept to relativelymodest levels.

An external subsidy for operating and maintenancecosts is widely regarded as inconsistent with efficientand sustainable provision of service.The private sectorgenerally perceives that a high level of risk is associatedwith external subsidy because of the possibility thatsupport will be withdrawn in the future. Therefore,relying on an external operating subsidy is also likely todiscourage private operators from taking part in PSPschemes.

The case for up-front capital subsidies, wherenecessary, is much stronger. Capital subsidies may en-able the extension of service to poorer groups thatcannot afford the full costs of connection but are ableto pay tariffs that cover operating and maintenancecosts.The private sector is likely to be less concernedabout a subsidy that is committed up front or at leastover a relatively short initial period of the project. Inaddition, the ability of customers to pay tariffs thatcover operating expenses is consistent with long-term sustainability of providing services. Until re-cently, the principal means of providing capital subsi-dies have been through cofinancing of concessionarrangements. Under this approach, one or more in-ternational lending agencies, usually through thehost government or a parastatal agency, will provide apackage of grant or soft-loan financing to the projectcompany.The company will then apply the grant orloan, alongside funds provided by its parents orshareholders and the commercial banks, toward cap-ital projects that serve both poor and relativelywealthy customers.

The main drawback associated with cofinancing isthe weak incentives for the operator to actually achieve


14

Table 1.1 Foreign Direct Investment in Rwanda

Year Amount (US$)

1979 13,000,0001989 16,000,0001998 7,000,0001999 2,000,0002000 8,000,0002001 5,000,0002002 5,000,000Source: World Bank 2003.

the level of performance—for example, utility connec-tions to poor households—that underlies the financialaid provided.Alternative forms of support that involvethe disbursement of subsidized financing only as andwhen there is tangible achievement against measurabletargets are currently being seen as a preferable alterna-tive.This performance-related approach to the provi-sion of aid funding is generally termed output-basedaid (OBA). OBA can be applied in a wide variety of

ways, though it is not appropriate to examine this sub-ject further in this report.This said, the use of OBA toprovide clear incentives to private operators to servepoorer communities should have important applicabil-ity in Rwanda. Its use should be an important consid-eration in relation to many of the potential PSPschemes identified in the CFR and may contribute toefficiency of provision as well as promote a reductionin poverty.

Introduction: Policies, Objectives, and the Role of the Private Sector

15

As a landlocked country with mountainous terrain,Rwanda faces unusually difficult problems in relationto both national and international transportation. Inad-equacies in its transport infrastructure impose signifi-cant additional costs on the country’s economy andrepresent a serious impediment to improving percapita incomes. Because of such inadequacies, thepoorest rural communities face considerable problemsgaining access to markets for their produce. Inferiortransport infrastructure—particularly the inadequaterural roads—also affects the ability of the poor to gainaccess to key services, such as health and education, andrepresents a serious obstacle to alleviating poverty.Moreover, deficiencies in Rwanda’s transport systemsspill over into the other infrastructure sectors, addingto maintenance costs and reducing the quality ofservice.

This chapter first considers the current organiza-tion and performance of the road subsector and setsout recommendations concerning how the subsectormight be improved, with particular reference to therole of private sector participation (PSP) in thisprocess. It then addresses the airport subsector, beforeexamining the key issues raised in connection withthe potential development of a railway infrastructurein Rwanda and the manner in which a railway de-velopment project might best proceed. The chapterconcludes by considering the potential to developwater transport, with particular reference to LakeKivu.

Roads

Overall responsibility for roads policy rests with theDirectorate of Roads within the Transport Departmentof the Ministry of Infrastructure. Until recently, theRoads Department, which was part of the Ministry ofInfrastructure (and formerly was part of the Ministry ofTransport, Public Works, and Communications), helddirect responsibility for maintaining all main roads and,it is understood, possessed the necessary equipment tocarry out this task.1 Under the government’s decentral-ization policy, responsibility for maintaining the un-paved national road and communal road network man-agement is, however, being transferred from centralgovernment to the provincial governments. Mainte-nance of paved roads is now carried out by the privatesector under national competitive tender arrangementsadministered by the National Tender Board, with su-pervision of road maintenance works carried out bythe private sector.Local communities undertake simplepaved road maintenance activities, such as clearingverges and drains, under the supervision of the Direc-torate of Roads.

The entire road network is state owned. Fundingfor road maintenance is provided through the RoadsMaintenance Fund (RMF), which was established as aroad fund about 12 years ago but was made an inde-pendent body in 1996. Income to the fund is derivedvariously from a charge on fuel, a toll on foreignvehicles, an axle-load charge, penalties for vehicle

16

2 The Transport Sector

The Transport Sector

17

overloading, and road damage compensation, plus, inprinciple, a government contribution.

The overall allocation of roles and responsibilitieswithin the roads sector is summarized in figure 2.1.

Policymaking, Planning, and RegulationAlthough the overall policy framework in transport iscurrently under review, it is understood that these arethe government’s policy objectives with respect to roadtransportation:• To enhance Rwanda’s integration into the regional

economy and to make Rwanda a regional trade andtransit center

• To focus transport sector investment on expandingand improving Rwanda’s infrastructure, protectingexisting capital investments, and improving roadsafety

• To institute a policy framework for the accelerateddevelopment of the road subsector

• To have road works contracts awarded throughthe National Tender Board, in collaboration withthe Ministry of Infrastructure and the Kigali CityCouncil

• To finance road maintenance works through theRMF, which is funded through the budget, a directlevy on fuel, a cross-border charge, and variouspenalty charges

• To encourage community participation in roadmaintenance through the district developmentcommittees

• To improve the availability and quality of local roadinfrastructure, thereby enabling the rural commu-nity to market its crops

• To create an environment conducive to the encour-agement of PSP in rehabilitating, maintaining, anddeveloping roads infrastructure.A number of measures have been taken or are

being introduced to implement these policy

Policymaking

Planning

Maintenance

Ownership

Funding

Operation

Centralgovernment

Regionalauthorities

Corporatesector

Community andindividuals

Public sector Private sector

Local communities are now providingsome road maintenance services onunpaved roads under the supervisionof regional road engineers.

Overall policy responsibility restswith Ministry of Infrastructure,Directorate of Roads.

The government of Rwanda hasownership.

Responsibility for funding rests withthe Roads Maintenance Fund, withthe shortfall, in principle though notin practice, being made up bygovernment. Some charges arecollected by agencies, such as theRevenue Authority and the TrafficPolice

Road works contracts to beawarded through the NationalTender Board, in collaboration withthe Ministry of Infrastructure andthe Kigali City Council.

Overall policy responsibility rests with Ministry of Infrastructure,Directorate of Roads, in liaison with regional authorities

Ministry of Infrastructure used to be responsible for maintaining all main roads. Maintenanceof the paved roads is now carried out by the private sector under national competitive tender arrangementshandled by the National Tender Board. The maintenance itself is supervised by the Directorate of Roads.

Figure 2.1 Roads: Roles and Responsibilities

Source: Adam Smith Institute research.


18

objectives, including the following:• The RMF has been established to finance road

maintenance and to strengthen the Directorate ofRoads. The incorporation of the Directorate intothe new Ministry of Infrastructure is also expectedto add to the department’s strength.

• Road works contracts are awarded through theNational Tender Board.

• In line with Rwanda’s broader decentralizationpolicy, responsibility for maintaining both the un-paved main roads and the unpaved rural road net-work is now being moved to the provinces and isbeing carried out by the local communities. Super-visory responsibility is to rest with the regional roadengineers of the former Roads Department, whoare being transferred to the various provincialgovernments.

Condition and Performance of the Road SubsectorThis section sets out available information on the ex-tent, condition, and performance of road assets. In ad-dition to information provided directly by the Min-istry of Infrastructure, which is used as the base data forthis Country Framework Report (CFR), the principalsources used were Scetauroute’s report on roadsprepared for the European Union (EU) in January2002 and relevant sections of the report titled “TheRwandan Economy: A Strategy for Investment,”which was prepared for the Rwanda InvestmentPromotion Agency (RIPA) in mid-2002 by the AfricaInstitute for Policy Analysis and Economic Integration(AIPA) of South Africa.The latter report appears to bebased on a review of other transportation reports dat-ing from 1997 to 1998.

Extent of the Road Network The total road networklength is about 14,000 kilometers, of which some5,350 to 5,408 kilometers (depending on source—seetable 2.1) constitute the classified main road network.2

The coverage of the network is said to be adequate.Theextent of the paved main road network is on the orderof 1,000 kilometers, although estimates again varyaccording to source (see table 2.1).The unpaved mainroad network extends some 4,300 to 4,400 kilome-ters, with exact estimates again varying, as shown intable 2.1.The paved road network connects the capitalto the main regional centers, as well as linking some of

the important regional centers to each other. It alsoprovides thoroughfares for corridor transport betweenBurundi, the Democratic Republic of Congo,Tanzania, and Uganda.

The balance of the road network amounts to about8,000 kilometers of unclassified roads, comprising over6,500 kilometers of unpaved rural roads and betweenaround 900 and 1,300 kilometers of urban roads (ap-proximately 500 kilometers of which are locatedwithin Kigali).

Records of the number of bridges, viaducts, andculverts are not available, although the number of suchassets is considerable because of the hilly nature of thecountry and the many rivers and streams that the roadnetwork has to cross.

Performance The condition of the classified road net-work was discussed in both the Scetauroute and AIPAreports and is summarized in table 2.2. (The Scetau-route report is thought to be the more accurate of thetwo reports.). The poor condition of the main roadnetwork can be attributed to the extensive damagecaused by unusually high rainfall in recent years, as wellas to a lack of equipment and the absence of both reg-ulations and enforcement measures to control the axleloads of heavy vehicles at the borders. It is understood,however, that a vehicle inspection center has beenbuilt, although it is not yet in operation, and that eightweigh bridges have been purchased but not yet in-stalled. It is essential that those facilities be put intooperation as quickly as possible.The return on this in-vestment should be rapid—especially if higher penalty

Table 2.1 Extent of the Classified Road Network(Kilometers)

Data Source

Ministry ofRIPA report EU report by Infrastructure

by AIPA Scetauroute dataType of road (mid-2002) (January 2002) (December 2002)

Paved main 930 1,022 1,100roads

Unpaved main 4,436 4,386 4,250roads

Unengineered 1,750gravel roads

Total 5,366 5,408 5,350Source: Data from AIPA 2002, Scetauroute 2002, and Ministry of Infrastructure.


19

charges are introduced, as recommended later in thisreport.

Maintenance carried out on unclassified roads hasbeen concentrated on the paved road system and ruralroads have suffered especially as a consequence. Theeffects have been exacerbated by Rwanda’s terrain andthe prevalence of narrow curves, steep gradients, anduneven surfaces, combined with heavy rainfall and poorsoil conditions. As a result, the condition of all unpavedroads, and especially unclassified rural roads, has deteri-orated to a relatively poor level. Rutting and potholesare common.

The state of the urban roads, especially in somedensely populated, hilly areas of Kigali, is very poorowing to heavy traffic, inadequate maintenance, andlack of proper drainage.This poor condition has led tounacceptably high numbers of accidents in the capital.The situation in Kigali is aggravated by the fact that the

utilities network is unmapped.A recent attempt to re-habilitate the roads in Kigali had to be suspended be-cause of the damage and disruption to water and tele-phone networks it caused.

A direct effect of the generally poor condition ofthe road network in Rwanda is the level of vehicle op-erating costs (VOCs), which is high by comparisonwith neighboring African countries—and the associ-ated reduction in the operational life of vehicles.Thesecosts represent a significant burden on the nationaleconomy. The World Bank (2000, p. 9) found that thelack of maintenance on the road between Gitaramaand Kibuye raised VOCs from a 1989 level of US$1.00per kilometer to almost US$3.40 per kilometer in1996.

In summary, the condition of the entire road net-work has deteriorated substantially in the past decade.This deterioration has been especially marked onunpaved classified roads and both urban and rural un-classified roads, resulting in a significant cost to theeconomy. As far as can be ascertained, the condition ofbridges, viaducts, and culverts mirrors that of the roadsof which they form part.

Impact of Poor Road Maintenance The current poor con-dition of the road network is damaging to Rwanda’seconomy because it restricts access to markets, raisesVOCs (making exports uncompetitive and adding tothe cost of imports), and reduces income from tourism.These direct effects on economic development in-evitably also present obstacles to reducing poverty andimproving the quality of life of a wide spectrum of thecommunity, most especially the most disadvantaged insociety.

Poor maintenance of the feeder roads, especiallythose in agricultural areas, has impaired market accessand hence restricted farming income for the ruralcommunity. It thus represents an obstacle to effectivereduction of poverty. In its recent review of the Trans-port Sector Project, the World Bank notes that rehabil-itation of the road between Gitarama and Kibuye hasresulted in VOCs being reduced by 50 percent,amounting to a reduction in overall transportationcosts of about 40 percent. As a result, agricultural sur-pluses in the area can now be sold in markets through-out the country, and a general shift is taking place fromsubsistence agriculture to production for the market.

Table 2.2 Condition of the Classified Road Network(percentage)

Data source

EU report RIPA reportby Scetauroute by AIPA

Category of road (mid-2002) (January 2002)

Category 1: international orcross-border roads

Very good and good 23 45Acceptable 37Fair 30Mediocre 30Poor 35Bad or very bad 10

Total 100 110a

Category 2: national roadsVery good and good 5Acceptable 20Undetermined 75

Total 100Category 3: communal roads,

including feeder roadsVery good and good 2Acceptable 8Undetermined 90

Total 100Categories 2 and 3Good 10Fair 40Poor 50

Total 100a.Total is as set out in AIPA 2002.Source: AIPA 2002 and Scetauroute 2002.


20

Opportunities for Private Sector ParticipationThe road rehabilitation, maintenance, and develop-ment program offers significant opportunities for PSP.PSP will need to be accomplished on a phased basis, toallow the private sector to gain confidence in operatingin Rwanda.The first phase should be to pass responsi-bility for a significant proportion of road rehabilitationand maintenance to the private sector through 5- or10-year concession contracts. Because it will probablytake a 10-year program to bring the entire road net-work back to an acceptable standard, a 10-year contractwould save rebidding costs at year 5.There would becontractual safeguards in the form of minimum per-formance standards,with penalties for nonperformanceand the ultimate sanction of the cancellation of theconcession.Where the cost of the rehabilitation is un-usually high owing to the poor state of the road, it maybe preferable to go to a 15-year contract to ease thestrain on cash flow over the concession period.

The intention would be that all rehabilitation workbe completed within the first 10 years. After year 10,concessions would be for maintenance only, to main-tain the existing and rehabilitated road network at anacceptable minimum standard (monitored by theDirectorate of Roads). The private contractor wouldcharge a service fee on a monthly or quarterly basis,made up of the rehabilitation costs, maintenance costs,initial mobilization costs, and a profit margin, whichmight be partially taken as dividends through the localproject concessionaire. The risks transferred to theprivate sector would in principle be the following:• The design, construction, and cost of the rehabilita-

tion work• The cost—but not the inflation risk—and the qual-

ity of the annual maintenance work, against anagreed-upon scope of work and services over theconcession period

• The completion date of rehabilitation work• General commercial and legal risks.The concession approach would start the developmentof a road rehabilitation and maintenance market inRwanda. In parallel, it will be important to support thedevelopment of this market by stimulating the privatesector—preferably the indigenous private sector—toset up companies owning and hiring road plant andequipment. After five years, the policy should be tomove progressively to design, rehabilitate, finance, and

operate (DRFO) concession contracts, under whichthe private sector would begin to invest in and take re-sponsibility for financing concessions for all road main-tenance projects. DRFO is a variation on the design,build, finance, and operate (DBFO) contracting modelwidely used to finance new road development. UnderDRFO arrangements, the private sector takes over re-sponsibility for an existing road.Whereas in the DBFOmodel the private partner funds initial road design andconstruction and meets ongoing maintenance costs, inthe DRFO model the initial financial burden arisesfrom the need for rehabilitation expenditure. As withDBFO, a long-term maintenance obligation falls onthe private partners.The private partner derives someor all of its income from actual or shadow toll pay-ments, thereby accepting a proportion of demand risk.

A credible and attractive scheme to incorporatePSP in road maintenance and development must nec-essarily focus in the first instance on the followingfour steps: (a) development of a prioritized list of ac-tivities, (b) design of measures to provide inputs intothe RMF, (c) capacity building to establish contractingout capability within the government, and (d) intro-duction of a method of contracting that includes localparticipation. Other countries in Africa have at-tempted a similar approach, and lessons can be learnedfrom their experiences. For example, Ethiopia, al-though it is a much bigger country than Rwanda, hashad a similar background of internal strife and disrup-tion. The process of transforming road maintenance(and rehabilitation and construction, including designand supervision activities) in that country has beenconducted along the lines mentioned above. InEthiopia, it has taken five years of concentrated effortto get to the stage where routine maintenance is en-tirely funded by road fund levies.

The problem of developing Rwandan private sec-tor capabilities in road maintenance and construction,as well as in design and supervision, can be addressedby twinning local and international contractors andconsultants. However, the time frame for achievingmeaningful domestic participation in these areas willnormally be long and will depend on the skill levelsavailable in the country at the start of such a program.

The first step is therefore to attract foreign roadcontractors with the necessary expertise and experi-ence in paved and unpaved roads—if possible, gained in


21

Rwanda or a similar African context—to bid for roadrehabilitation and maintenance contracts, preferably ona 5- to 10-year concession basis. Attracting such con-tractors will require immediate review and reform ofthe law on foreign investment, the company law, andthe tax and import duty laws. It will also require thepreparation of a credible national and internationalpublicity program built around a clear, properly speci-fied and financed plan for road rehabilitation andmaintenance. It is vital to ensure that there is more thanone foreign road contractor available in Rwanda to bidfor road projects.

The second step is to encourage suitable local con-tractors to diversify into the road infrastructure sector,to acquire the necessary expertise, and to purchase thenecessary equipment.Rwanda has at least one medium-size local private building contractor capable of con-structing buildings of reasonable size and to a goodstandard. Contractors with a proven track record in asector such as building construction can be expected toperform well in the road sector.Apart from addressingthe cross-cutting legal issues, attracting local contractorswill again depend on a credible publicity program forthe road rehabilitation and maintenance plan,backed bytheWorld Bank and the donor community.

The third step is to attract local business peoplewith the necessary funds to invest in new, local roadcontracting companies.The new companies will needto recruit a staff with the necessary expertise, althoughat first they may prefer to hire plant and equipmentfrom other contractors. While building up their ex-pertise, locally based enterprises are likely to bid onlyfor the smaller rehabilitation and maintenance projectsrelating to unpaved roads, but this is an important be-ginning from which the indigenous road maintenancesector can develop. Local enterprises will bid only ifthe government publicity campaign, backed by theWorld Bank and the donor community, is convincingin promoting this opportunity.The ability of the pri-vate sector to react to opportunities where it is con-vinced that a viable new market is opening up has beenwidely demonstrated worldwide.

A further step would be to attract local businesspeople to set up new companies to provide the neces-sary road plant and equipment for these new road con-tractor companies on a rental or lease basis. Again,these local investors will need to be convinced by

government publicity that such investment is a new,viable market opportunity.

Given the size of the country, it is obviously notnecessary to establish a large number of foreign andlocal road contracting companies or local plant andequipment hire companies.Two or three hire compa-nies, three foreign contractors, and perhaps four or fivelocal contractors would provide adequate supply-sidecapacity and competition. Apart from the obviouseconomic benefit of an improved road network, the es-tablishment of such new companies would have theadded economic benefit of increasing employment op-portunities, particularly in urban areas, with a resultantboost to the poverty reduction program.

Involving local communities in maintaining therural road network, with the right controls on quality,makes a great deal of sense. The promotion of labor-intensive public works is an important plank of broadergovernment policy. In some areas, local communities,encouraged by the local prefects, have organized roadmaintenance gangs and have carried out road mainte-nance on unpaved roads in their area under the super-vision of the Directorate of Roads.These local com-munities can be encouraged to set up propercommercial companies for maintaining rural roads.They will require assistance from the government and,in particular, the Directorate of Roads for training toestablish the necessary level of expertise.Once again, inaddition to its direct benefits, such a program wouldalso bring indirect economic benefits in terms of in-creased rural employment opportunities and be a steptoward reducing poverty.

IssuesSeveral issues need to be addressed to move for-ward with the road rehabilitation and maintenanceprogram.

Maintenance of the Road Network The maintenance ofthe entire road network is currently inadequate and isconstrained by the lack of road maintenance funds. InAugust 2002, the government of Rwanda adopted adocument defining strategic prioritization for theroads. The document stated that resources should bedistributed to national roads of international interest,national roads of national interest, roads of local (com-munal) usage, and the urban road network. However,


22

the implementation of this policy remains question-able. The government must still go some way if scarceresources are to be applied where they will generatethe greatest contribution to achieving the govern-ment’s broader economic and social objectives. Thereis no road management system in Rwanda and no sys-tematic planned road maintenance program appears tobe in place. Instead, maintenance is carried out ad hocor in response to urgent needs for repairs to the pavedroad network. There is a clear need to develop a sys-tematic, planned approach to rehabilitation and main-tenance, particularly if the private sector is to be effec-tively involved in the execution of such a program.This road rehabilitation and maintenance plan willrequire the backing of the World Bank and the donorcommunity, supported by their commitment in princi-ple to provide appropriate financing.

A necessary first step toward developing a plannedprogram is prioritization of roads in order of their im-portance to the economy.The following order is sug-gested:1. Major export-import routes between the capital

and the Tanzania and Uganda borders2. Roads between Kigali, Ruhengeri, and Gisenyi 3. Roads between Kigali, Butare, Cyangugu, and

Bugarama4. Paved roads linking Rwanda to Burundi and the

Democratic Republic of Congo5. Feeder roads from major agricultural areas to the

main export roads6. All other unpaved main roads7. All urban roads8. Rural roads leading to important tourist sites such

as national parks9. All gravel roads

10. All other rural roads.A further problem is the scarcity of experienced

road contractors in the country. As of the date of thisreport, only one foreign-owned contractor was presentin Rwanda, a Chinese company. A second contractor,Astaldi of Italy, had recently pulled out of the country,apparently in response to the low level of ongoing roadconstruction and maintenance work. There were,furthermore, only four local contractors, two of whichhad only limited equipment and two of which had noequipment whatsoever. Equipment requirements weretherefore largely met by the Chinese company. The

four Rwandan contractors also lacked both expertisein and experience with tarmac surfaces.

A direct consequence of the current decentraliza-tion policy is likely to be that overall unpaved roadmaintenance activity—and quality—will become verydifficult to control.The Directorate of Roads is tryingto counteract this effect by providing training to localcommunities. The department is also consideringhiring consultants directly to take responsibility for allroad maintenance on a regional basis. Whether thisapproach will be approved and implemented is not yetclear.Nor is it evident how the approach will be fundedor whether it will in fact lead to any improvement inthe current level and quality of road maintenance.

As regards unpaved gravel roads, the AIPA reportremarks that the 15 percent saving in construction costfor an unpaved road by comparison with the cost of anequivalent paved road is outweighed by the ongoingmaintenance costs of unpaved roads, which may be upto 10 times those of paved roads.As a result of gener-ally inadequate expenditure on road maintenance,these roads have deteriorated to such an extent thatthey are now in urgent need of rehabilitation.

It is clear that urgent steps need to be taken to im-prove the level and quality of road maintenance acrossthe whole network. The objectives must be both torestore the network to an acceptable standard and toprevent further deterioration beyond a point thatwould entail vastly more extensive and costly repairsand reconstruction in the future.

It is not clear how environmental considerations arecurrently taken into account or will be taken into ac-count in any future road rehabilitation or maintenanceprogram.

Roads Maintenance Fund Although revenue collectionhas recently increased slightly owing probably to ahigher level of penalties imposed or a greater numberof foreign trucks in transit—no government contribu-tion has been paid to the RMF throughout the fund’sexistence. The Revenue Authority and the TrafficPolice, who are responsible for collecting some of thecomponent charges, have also been very slow in turn-ing over to the RMF the amounts they collect; bothhave been persistently in arrears. Combined with fail-ure to enforce charges, these delays have resulted inan annual collection level that has fallen to about


23

US$4 million currently from about US$10 million be-fore 1994, according to the AIPA report. Furthermore,charge levels were set 12 years ago and have not beenincreased in line with inflation over this period. TheRMF has made a request to the government for anurgent increase, so far without result.

The overall outcome is that the RMF can providefunding for only 16 percent of the FRw 12 billion (ap-proximately US$24 million) of annual maintenancespending requests, although this level is itself an inade-quate reflection of actual spending needs. As a result,the total maintenance backlog is growing year by year,and the road network continues to deteriorate. Urgentsteps are therefore required to make the operation ofthe RMF effective.

Estimated Cost of Road and Bridge Rehabilitation and Main-tenance An indication of the potential cost of a road re-habilitation program can be found in the Scetaurouteand AIPA reports. The principal relevant findings ofthese studies are set out in appendix C of the CFR. Insummary, the AIPA study suggests rehabilitation andmaintenance costs over 10 years of US$186 million forthe paved road network, US$253 million for the un-paved main road network, and US$395 million for theunpaved secondary road network—US$834 million intotal. The Scetauroute report suggests figures rang-ing very approximately from US$130 million toUS$180 million in total over a 10-year period, coveringonly the paved and unpaved classified roads system.Although the Scetauroute report covers a smaller pro-portion of the total road network, the equivalent costidentified is clearly much lower than in the AIPA study.There is no basis on which it is possible to judge thereliability of the different cost estimates.

Financing of Road Rehabilitation and Maintenance Thegenerally adopted approach is for rehabilitation andmaintenance to be financed by road users. This ap-proach can be difficult to apply in developing coun-tries with low traffic levels and low income levels, al-though car and commercial vehicle owners are usuallyin the higher income brackets. Those with lowerincomes—who will normally travel on foot or by bus,animal-drawn cart, or bicycle—will either face no roaduser charges whatsoever or benefit from the dilution ofcharges among a group of passengers. Nevertheless,

direct road-user charging (through tolls) is usually un-popular with both politicians and the general public.The cost of providing road infrastructure is thereforenormally financed from annual vehicle registrationcharges, with the balance met from fuel taxes and gen-eral taxation.

In any case,until the private sector gains experienceand confidence in operating road rehabilitation andmaintenance concessions in Rwanda, it will necessarilyfall to the government to raise the required financing.The following approaches to financing would thus ap-pear possible for a 10-year program:• Increase RMF charges and penalties to a level that

would finance the fully costed plan on an annualbasis over the 10-year time frame—with chargesand penalties subsequently to be increased annuallyin line with inflation. If the government view is thatthose charges would be excessive, then a smaller ac-ceptable increase in charges might be introduced(although penalties might still be increased by thefull amount), again indexed to inflation, with theshortfall to be financed from general taxation orborrowing.

• Impose direct road-user charges in the form ofroad vehicle tolls, which would be charged andcollected by the government on the paved roadnetwork only or, alternatively, on paved roadsbeing rehabilitated or maintained by the privatesector on a concession basis (with the private sec-tor responsible for the toll collection). In view ofconcerns regarding low traffic flow and affordabil-ity, the initial level of tolls could be low. The levelwould gradually increase as road traffic increasedand users became accustomed to the tolls. Tollswould have the short-term disadvantage of increas-ing VOCs, although this effect would diminish asthe paved roads were brought back to an acceptablestandard. The initial cost of constructing the tollbarriers and ongoing costs of toll collection wouldalso be a drawback of this approach.3 Potential pri-vate sector participants could also well perceive asignificant risk that the income from affordabletolls would fall short of the amount required tomeet the concessionaire’s service charge, giventhe current, relatively low traffic flow on manyRwandan roads. The private sector may be pre-pared to become involved only if there is a clear


24

and credible commitment by the government tomake up any shortfall from taxation or borrowing.

• Impose indirect road-user charges through “shadowtolls”using roadside meters,which would count thepassing traffic, broken down by vehicle length, andthen calculate the total shadow revenue at theagreed-upon shadow tolls for the different vehiclelengths. Payment of charges would be made by thegovernment from general taxation and borrowingbecause there is no direct monetary payment byroad users. Again, these tolls could be applied to thepaved road network as a whole or to paved roadsbeing rehabilitated or maintained by the privatesector under concession agreements. Shadow tollsare generally popular with politicians because theyavoid publicly unpopular, directly collected tolls.

• After year 5—when the private sector has gainedexperience and, it is to be hoped, confidence in op-erating concessions in Rwanda—it may be possibleto introduce a rehabilitation-focused variant of theDBFO concession contract model. Such conces-sions would transfer the risk of raising the financingfor these contracts from the government to the pri-vate sector. It may be necessary to increase the ma-turity of the DBFO contract to 15 years, to ease thestrain on the cash flow in the early years.

• Use a combination of the above options, so thateach option is seen neither as too severe or exces-sive nor as premature in the context of a newly de-veloping market. In combination, these options aresufficient to raise the level of financing required, aswell as being acceptable to the government androad users. The financing of a road rehabilitationand maintenance program will require strong, co-ordinated support from the World Bank and thedonor community, particularly in regard to con-tributing toward the financing costs to be borne bythe government.

Financing of Essential New Road Development The exist-ing rehabilitation and maintenance program shouldtake clear precedence over new road constructionthroughout the next 10 years, in terms of allocation offinancing and other scarce resources. The questionthen is how essential new roads or bridges should be fi-nanced over this period. A vital first step will be a clearassessment of what limited construction is necessary

over the period, taking into account the rehabilitationprogram. Any study of the need for new roads shouldbe driven and constrained by a requirement for consis-tency with the new transport sector policy, which wasbeing prepared with the assistance of the World Bank’sSub-Saharan African Transport Policy Programme atthe time this report was written.

Again, there are several financing options:• Impose a five-year general moratorium on all new

road construction, after which new projects wouldbe advanced only through public procurement, as-suming the necessary capital financing can beraised. Exceptions would be allowed only in specialcircumstances (for example, to meet road infra-structure needs for the resettlement of returningrefugees).

• After the same five-year moratorium, proceed withnew schemes only on the basis of private procure-ment on a DBFO concession basis, under whichthe private sector would have responsibility for rais-ing financing.

• Proceed with a construction program for essentialnew roads—as identified by the Directorate ofRoads in the new roads study—by public procure-ment only for the next five years. This approachwill give the private sector time to gain confidencein the government’s rehabilitation and maintenanceprogram, in the new market for road rehabilitationand maintenance, and in operating in Rwandaunder concessions.After year 5, switch to a privateDBFO concession basis.The feasibility of this ap-proach would depend on the government’s abilityto raise the capital financing necessary to build newroads during the initial five-year period, on top ofthe financing for the rehabilitation and mainte-nance program.

Institutional Capacity The ability of the Ministry of In-frastructure to bear the burden of addressing the widerange of issues arising in the roads sector is potentiallya matter of some significance.The question of overallinstitutional capacity is considered elsewhere in this re-port. It is clear, however, that addressing these issueswill require considerable effort, involving hiring expe-rienced personnel; reviewing current departmentalwage scales in order to attract and retain the necessarypersonnel; and carrying out a continuing in-house


25

training program, together with donor technical assis-tance programs. Such a training program is, however,essential if the government’s objective of attracting pri-vate sector involvement in the provision of public in-frastructure services is to succeed. Such a program willalso need the full support of the World Bank and thedonor community.

An approach that has been used in a number ofcountries is to transfer responsibility for the executionof roads policy to a roads agency operating at arm’slength from the government, which continues to beresponsible for policy development. The benefits ofsuch an approach in Rwanda would be to transfer tothe new agency, the responsibility for performing allroad rehabilitation, maintenance, and construction, to-gether with the responsibility for establishing and en-forcing design and maintenance standards. It wouldalso free the Directorate of Roads from ministry bu-reaucracy and allow it to concentrate on improving theefficiency and performance of road rehabilitation,maintenance, and construction. The new agencywould also take over responsibility for all dealings withthe private sector related to participation in providingroad infrastructure services, within the policies set bythe Ministry of Infrastructure.

RecommendationsThe following are the CFR recommendations:• Steps should be taken, in cooperation with the

provincial governments, to return responsibility forall road and bridge maintenance to the Directorateof Roads of the Ministry of Infrastructure in orderto reestablish capacity to undertake road mainte-nance and to improve quality.

• The Directorate of Roads should quickly prepare aminimum set of design and maintenance standardsfor paved, gravel, and unpaved roads, as well as forbridges, viaducts, and culverts. The departmentshould subsequently be held responsible for main-taining these standards at all times.

• The Directorate of Roads should draw up a set ofenvironmental guidelines to be applied to all roadand bridge construction, rehabilitation, and mainte-nance programs and should be held responsible forenforcement of these guidelines.

• The Directorate of Roads should prioritize roads inorder of their importance to the economy and

should give rehabilitation and maintenance prece-dence over new road construction, unless there is avery strong economic case for the new road.

• Based on the road priority list, the Directorate ofRoads should establish a fully costed plan of roadand bridge rehabilitation and maintenance cover-ing a 10-year period, with a scope that embracesthe entire road network.

• As a matter of urgency, axle-load restrictions shouldbe enforced against overloaded vehicles to preventfurther damage, in particular to the paved road net-work. This task may well require additional re-sources being made available to the enforcementauthorities.

• Serious consideration should be given to introduc-ing penalties for vehicles that cause environmentaldamage to the road network and its rights of way, asa consequence of accidents or technical failures. Inaddition, the law should be strengthened to protectand enforce rights of way.

• There should be an immediate increase in all RMFcharges to take account of inflation over the past12 years. The charges should then be further re-viewed in light of their adequacy to meet the re-quired level of annual maintenance funding.

• The government should make up any shortfall eachyear in the RMF against the level of fundingrequired to bring the road network back to an ac-ceptable level over an agreed-on timescale.

• The collection authorities for certain charges(namely, the Revenue Authority and the Traffic Po-lice) should be held responsible for passing on thesecollected charges within three months of receipt orpaying penalty interest.

• Immediate action should be taken to address insti-tutional capacity problems within the Directorateof Roads:— To prepare it to handle its increased responsibil-

ities in controlling all road and bridge mainte-nance

— To enable it to establish and manage the in-creased participation of the private sector in theroad infrastructure sector and to meet the asso-ciated contract monitoring requirements

• The government should give serious considerationto establishing the Directorate of Roads as an inde-pendent roads or highways agency, reporting to the


26

Ministry of Infrastructure, with the Ministry of In-frastructure retaining responsibility for roads policy.

• To attract the private sector to participate in roadrehabilitation and to establish a road maintenancemarket,with participation by both foreign and localprivate sector entities, including the local commu-nities, the government should move all rehabilita-tion and maintenance contracts to a 10-year con-cession basis. There will be contractual safeguardsfor the government through minimum perform-ance standards with penalties for nonperformanceand the ultimate sanction of cancellation of theconcession.

• A financing plan should be agreed on with donorsfor the 10-year road rehabilitation plan. It shouldinclude consideration of the following options:— The RMF— Government borrowing with the support of the

donor community— Road tolls—direct or indirect (shadow) on a

staged basis within affordability levels

— In five years, the possibility of moving the con-cession contracts onto an appropriate variant ofDBFO.

Airports

Key roles and responsibilities within the airport sectorare summarized in figure 2.2.

Policymaking, Planning, and RegulationThe government’s objectives for airports are• To assist the development of air transport• To enhance Rwanda’s integration into the regional

economy• To make Kanombe Airport a regional transport

hub and part of a proposed export-processing zoneA number of measures have been taken to implementthe policies and objectives:• Kanombe Airport in Kigali is shortly to be up-

graded with a repaved and slightly extended run-way, a new taxiway, an increased aircraft parking

Policymaking

Planning

Regulation

Ownership

Funding

Operation

Centralgovernment

Independentagencies Parastatals

Corporatesector



Planning is carried out bythe state-owned AirportsAuthority, with itsheadquarters at Kanombe.

All the airports andairfields are operatedand maintained by theAirports Authority.

Catering (limited) and carparking are operated bythe corporate sector.

The government of Rwandahas ownership.

The Ministry of Financeand Planning is responsiblefor funding.

Overall policy responsibilityrests with Ministry ofInfrastructure, Departmentof Transport

Figure 2.2 Airports: Roles and Responsibilities



27

area,new navigational aids,new fire-fighting equip-ment, and a renovated lighting system.

• Kamembe Airport, near Cyangugu, is being reha-bilitated with a new aircraft parking area and therenovation of the existing buildings.

Sector PerformanceThere are two airports and five airfields in Rwanda.The one international airport is Kanombe Airport inKigali, which has a paved runway of 3.5 kilometersand handles about 140,000 passengers annually, plusabout 6,500 tons of cargo (as of 2001). The relativelynew terminal is designed to handle up to 500,000 pas-sengers annually, so it has ample spare capacity at pres-ent. The second airport is the domestic airport atKamembe near the port of Cyangugu on Lake Kivu inthe southwest. It has a paved runway of 1.5 kilometersand handles some passenger and cargo traffic emanat-ing from the Democratic Republic of Congo. Theonly other airfields with paved runways are at Gisenyi(1 kilometer), which is also on Lake Kivu but in thenorthwest, and at Butare (less than 1 kilometer),which is in the south.The other three airfields are atRuhengeri in the north, Nemba in the south, andGabiro (a military field) in the east. All the airportsand airfields are operated and maintained by the state-owned Airports Authority, with its headquarters atKanombe Airport.

Kigali is served by one European airline (SNBrussels), four African airlines, and one local airline.There is also one local and one regional regular cargoairline service and occasional general cargo flights, in-cluding occasional Russian cargo planes to Kanombeand Kamembe. In addition, there are two small domes-tic charter plane companies. Kamembe receives tworegular DHL courier planes a month. The other air-fields are used infrequently.

As with road transport, the cost of air transport ishigh, owing to lack of competition and the small sizeof the market. Air cargo rates for general cargo fromKigali to Brussels and London are set out in table 2.3 asa benchmark.

Issues

Development Priorities The following issues emergedduring the development of the CFR:

• The need at Kanombe for increased export-importcapacity, with bigger refrigeration storage and alarger general storage area

• The requirement for a new taxiway at Kanombeand an extension to the apron for parking cargoplanes

• The requirement for a new emergency pavedairstrip for Kigali, in case Kanombe is closed (a sitehas been identified at Bugasera in the Ngendaregion)

• The need for rehabilitation of the runway andapron at Kamembe Airport

• The need for rehabilitation of Gisenyi andRuhengeri airstrips should tourism increase topre–civil war levels

• The need to consider a future international airportlocation, such as the site at Bugasera, given thatKanombe cannot be developed further owing tothe difficulties of the site.The need for a new emergency paved airstrip for

Kigali and the potential need at some time in thefuture for a new international airport to replaceKanombe lie beyond the scope of the CFR and willrequire extended technical and economic analysis.

Opportunities for Private Sector ParticipationThere seems little potential for the private sector toparticipate in this sector in the short term. A numberof options should be considered in the medium term,including the introduction of a private sector contractto manage airports.The contract would be monitoredby the Airports Authority, which could itself be priva-tized if the economy develops and passenger numbersgrow. In the short term, it may also be worth consider-ing the use of PSP options on a smaller scale—for

Table 2.3 Air Cargo Rates for General Cargo

SN Brussels Kenya Airways

Rate RateSize (kg) (US$ per kg) Size (kg) (US$ per kg)

0–44 6.95 0–99 2.7045–99 5.50 100–499 2.55

100–499 2.45 500–1,000 2.25500–1,000 2.35Sources: Data from SN Brussels and Kenya Airways.


28

example, contracting out airport car parking to thelocal private sector.

RecommendationsThe following are the CFR recommendations:• Investigate the need for an increase in the refriger-

ation and general storage facilities, the case for anew taxiway, and the case for extension of theapron in the context of the upgrading project forKanombe Airport. If justified, these project compo-nents should be included in the project investmentprogram.The prime objective would be to ensureno constraint on the growth of passenger and cargotraffic, which is important because of its potentialfor generating increased foreign exchange revenue.

• Rehabilitate the runway and apron at KamembeAirport.

• At the appropriate time, conduct a proper study ofthe rehabilitation of Gisenyi and Ruhengeri air-fields.This study is unlikely to be justifiable in theshort term. It should be initiated only if tourism re-turns to pre–civil war levels and small air charterscan compete against road transport on an improvedroad network.

• Give serious consideration to establishing theAirports Authority as a civil aviation authority,which—in addition to its current responsibilities—would take over responsibility for all civil aircraftmovements within Rwandan air space.This changewould bring responsibility for all airports, air safety,and civil aircraft movements within Rwanda underone agency.

Rail

Policymaking, Planning, and RegulationThe government’s key policy objective for rail is to in-vestigate the feasibility of establishing a railway to con-nect through one of the other regional railway systemsto the port of Dar es Salaam or Mombasa. This linkcould bring a number of benefits:• A reduction in the transport costs of exports and

imports because of the ability of railways to carrygoods in bulk

• Lower wear and tear on Rwanda’s road networkand reduced maintenance and life-cycle costs

• Reduced dependence on the road networks ofKenya, Tanzania, and Uganda—which are not ingood condition or are unpaved.Given the prospects of increased trade, the govern-

ment is also keen to consider the possibility of estab-lishing a southern transport corridor linking thecountries of the Great Lakes Region with SouthAfrica, as part of the Great Lakes Railway Project.Thisproject uses a combination of rail and lake transport.Adiscussion document was prepared by Protekem andthe Common Market of Eastern and Southern Africain 2000 to explore the available options.

IssuesThe case for the proposed rail development requirescareful assessment, based on a full feasibility study and aproper environmental impact assessment (EIA). TheChinese government has apparently made an offer tofinance such a feasibility study, with obvious interest inthe prospects of building such a railway if it goes ahead(Economist Intelligence Unit 2000).

The main concern relates to traffic potential. It isalmost an axiom that a new railway line is financiallyfeasible or economically viable only if it has guaranteedthroughput of 1 million metric tons of paying freighttraffic on the length of track constructed and if theterrain is easy and rolling, does not involve steep gradi-ents, and does not require a large amount of bridgingand tunneling.Where the terrain and soil present diffi-culties, as is the case in Rwanda, one can easily expectthe threshold of viability to rise dramatically—perhapsto twice the usual rule of thumb. Considering theshort distances within Rwanda, an internal railwayline would not be expected to be viable at all, incomparison with a road, unless large amounts ofminerals will need to be moved from mining areas orrefineries.

There is a significant question about whether sucha large investment could be justified, given the currentand projected volume of Rwanda’s exports and im-ports.These doubts might be mitigated if the railwaycould be connected to Burundi and the DemocraticRepublic of Congo—in particular, to the mineral-richKivu provinces of the eastern Democratic Republic ofCongo, which have no outlet to the East African coastexcept by a long road route. A regional conceptshould also be more interesting and acceptable to the


29

donor community, if the economics can be made towork.

The difficult topography of Rwanda—with itsmountains, hills, and valleys—would make a railwayvery expensive to construct.This expense might be al-leviated to some extent by routing the railway fromIsaka in Tanzania, where it would connect to the Tan-zanian Railways line from Dar es Salaam to Mwanza,through Burundi to the south of Rwanda, approachingKigali from the south in the Ngenda and Gashora re-gion.This routing might also support a regional railwaysolution, but it would not resolve the need for a link byroad or rail to Lake Kivu to complete the connectionwith the Democratic Republic of Congo. Anotheroption would be to extend the railway throughBurundi around the south of Rwanda to connect withthe Butare road at Akanyaru Haut on the border.Tradewould then come by road through Cyangugu andButare to transfer to the railway.

An alternative to connecting to theTanzanian Rail-ways at Isaka (as investigated by Austria RailEngineering in 1984 in the “Kagera Basin RailwayStudy,” the executive study of which was published in1991) would be to build a new railway from KemondoBay in Tanzania, a port on the west shore of LakeVictoria.This line would run through Tanzania to theeast of Rwanda, entering Rwanda at Rusumo in thesoutheast for a distance of 4 kilometers, and then toBurundi to a terminus at Kabanga. Rwandan tradegoods would have to be carried by road to the railterminus at Rusumo. The advantage of this approachwas identified by its shorter track length—about270 kilometers—associated with a connection by boatto the port of Mwanza and the railway to Dar esSalaam, a connection by boat to the port of Kisumu inKenya and the railway to Mombasa, and a connectionby boat to the port of Jinga in Uganda and an alterna-tive rail connection to Mombasa.The port of KemondoBay would require upgrading to handle the increasedtrade.

All these routes would rely on the infrastructureand operating efficiency of Tanzanian, Ugandan, andKenyan Railways.The infrastructure of these railwaysis considered to be of variable quality and reliability,with their operating management less than efficient.In addition, the track capacity of these systems mighthave to be increased to handle the increased flow of

goods—and this improvement also would have to befinanced.

Although the potential project to establish a south-ern transport corridor linking the countries of theGreat Lakes Region with Southern African railwaysystems incorporates the necessary regional approach,it will be a very expensive project. Its economic andfinancial feasibility will need to be very thoroughlytested before it can be regarded as a project for seriousconsideration.

RecommendationsThe government is committed to pursuing the imple-mentation of a railway project and to seeking donorcommunity support for what will be a major enterprise.It will therefore be necessary to approach the donorcommunity for support in updating the 1991 KageraBasin Railway Study. This exercise will need to extendthe scope of the original study to include the Democ-ratic Republic of Congo, in addition to Rwanda andBurundi. It is likely that only a regional approach willbe acceptable to donors and, moreover, that the projectwill not be economically viable unless it includes theconsiderable mineral trade and transport of the Kivuprovinces of the Democratic Republic of Congo.

Should such a study be carried out and the projectprove to be economically viable, an EIA would alsobe necessary before financing could be considered.Because the government also wishes to pursue the al-ternative of the southern transport corridor project, itwill again be necessary to approach donors and theother governments involved and to carry out the nec-essary feasibility study and EIA to establish whether theproject might be viable and which project of the two isthe most economically and financially viable.

Water Transport

At the request of the Working Group, the CFR exer-cise was to include a review of three studies on thesubject of water transport:1. A study for the development of transport on Lake

Kivu, dated November 1986, prepared for the Eco-nomic Commission for Africa (ECA) of the UnitedNations


30

2. A preliminary study on the navigability of the RiverKagera, dated December 1986, also prepared for theECA

3. A study of the need for a boatyard on Lake Kivu bythe Bureau National d’Études de Projet (BUNEP),dated July 1986.

However, it did not prove possible to obtain access tothe River Kagera study, and this topic has been ex-cluded from the CFR.

Sector PerformanceTable 2.4 through table 2.6, together with the support-ing text summarize key relevant data extracted fromthe reports concerning water transport on Lake Kivu.

The BUNEP report also refers to the existence of2 passenger boats (motor launches) that can carry up to50 passengers each and 11 barges with a total carryingcapacity of 800 tons. It has not been possible to con-firm the existence of either the passenger boats or thebarges. Both cement and beer are still transported onthe lake, which indicates that some barges have sur-vived, although no passenger transport appears to beavailable.

Opportunities for Private Sector ParticipationThere appears to be the potential for the transport ofgoods on Lake Kivu on an economically viable basis.Examples would be the distribution of beer from thebrewery at Gisenyi to the port of Kibuye, where itcould be transferred to road transport for national dis-tribution, and to the port of Cyangugu for local distri-bution and possibly export to Burundi and the Demo-cratic Republic of Congo. Another example—in theopposite direction—would be the distribution ofcement from the manufacturing plant at Cyangugu tothe ports of Kibuye and Gisenyi, where it could betransferred to road transport for national distributionand possible export. Another example could be thetransportation of tea, which is grown near bothCyangugu and Gisenyi, to Kibuye for road transporta-tion to Kigali and for export.These possibilities wouldpotentially maximize the efficiencies of both water androad transport.

Another potential economically viable use of watertransport is to provide market access to the small-scalecoffee growers of Nyamyumba and Kayove districts ofGisenyi province. These districts front on Lake Kivuand have very poor rural road connections to theunpaved road between Gisenyi and Kibuye. In addi-tion, a number of coffee-washing plants are located ad-jacent to the lake.There is potential for the coffee to becollected by water transport from the coffee-washingplants and then transferred to road transport at eitherKibuye or Gisenyi. The same opportunity is likely toarise in the coffee-growing districts south of Kibuyetoward Cyangugu.

IssuesA major constraint on the potential to develop watertransport on Lake Kivu is the absence of a boatyard on

Table 2.4 Commercial Traffic

Travel route Weight

Gisenyi to Kibuye 13,500 tonsKibuye to Cyangugu 11,860 tons

Total 25,360 tons

Kibuye to Gisenyi 11,380 tonsCyangugu to Kibuye 6,200 tons

Total 17,580 tonsLake fleet:2 motor launches carrying 50 passengers each2 motor launches carrying 15 passengers each 13 barges with capacities from 10 to 110 tons 30 tugs of various sizes Source: ECA 1986.

Table 2.5 AnnualTraffic from Gisenyi to Cyangugu, 1985

Type of traffic By lake By road

Passengers 24,000 132,000Goods 33,600 tons 7,100 tonsGoods including Congo, 58,953 tons —

Dem. Rep. of

— Not available.Note: Goods comprise raw materials, cement, beer, and agricultural produce.Source: BUNEP 1986.

Table 2.6 Estimated Traffic, Gisenyi to Cyangugu, 1990

Type of traffic By lake By road

Passengers 55/65,000 230/235,000Goods 80,000 tons 15/30,000 tonsNote: Goods comprise raw materials, cement, beer, and agricultural produce.Source: BUNEP 1986


31

the Rwandan side of the lake with facilities either forboat and barge maintenance and repair or for newconstruction. Any new boats, tugs, or barges wouldhave to be imported overland from Mombasa or Dar esSalaam, which would be difficult and expensive. It isdifficult to see how water transport on Lake Kivu canbe further developed or, indeed, maintained at currentlevels over the medium term without at least one boat-yard available for boat, tug, and barge maintenance andrepair. Furthermore, at least one boatyard needs to havethe capacity to build new boats, tugs, and barges, with-out which it will not be economically viable to usewater transport on Lake Kivu because of the high costof importing such boats.

RecommendationsThe CFR recommends that the study for the develop-ment of transport on Lake Kivu be updated and aug-mented to include an update of the BUNEP report onthe construction of a boatyard on Lake Kivu.

Notes

1. The department still holds some maintenance equipment, butthe equipment is no longer being replaced.2. The classifications are as follows: category 1—internationalor cross-border roads, category 2—national roads, and category3—communal roads including feeder roads.3. Note that in the alternative case these costs would be included inthe private sector’s costs to be recovered from the tolls.

This chapter describes the current structure of the en-ergy sector and provides a synopsis of existing govern-ment policy with respect to energy services in general,rural electrification in particular, and other key policyelements such as sector regulation and environmentalstrategy.1 It further summarizes sector performance interms of the character and quality of service delivery tocustomers, the financial state of the industry, and thenature and condition of existing assets. Opportunitiesfor private sector participation (PSP) in Rwanda’s elec-tricity and natural gas subsectors are identified alongwith key sectoral issues. The section concludes withrecommendations for improving the environment forPSP in the energy sector.

Sector Structure, Roles, and Responsibilities

Figure 3.1 illustrates how the various elements and in-stitutions involved in energy sector policy and servicedelivery are expected to evolve over the coming years.The provision of electricity (and water services) in theurban districts of Rwanda is the responsibility of Elec-trogaz (ELG).The key features illustrated in figure 3.1concern the transfer of ELG’s operations to a manage-ment contractor in March 2003 (depending on negoti-ations), the potential for ELG subsequently to be put ona concession contract or even fully privatized, the likelyintroduction of independent power producers into thesector, and the transfer of some elements of industry su-pervision to a Multisector Regulatory Agency (MSR).

Policymaking, Planning, and Regulation

The government of Rwanda—particularly the Min-istry of Infrastructure’s Energy Division (MIED)—hasnot developed and published a single comprehensive,coordinated policy statement and program for theenergy sector. It has, however, prepared a general strat-egy for reforming and developing the sector, bothwithin the context of the ongoing PSP initiative withrespect to ELG and as part of its poverty reductionstrategy policy (PRSP).This policy statement needs tobe published as soon as possible to help shape thedevelopment of the sector.

The government recognizes that to promote eco-nomic development, it must increase access to energyin rural areas. Doing so can add to off-farm employ-ment opportunities in, for example, agroprocessing andother small-scale manufacturing enterprises that canhave a direct effect on the reduction of poverty.Clearly,the quality of life for rural residents will also be im-proved through domestic access to, and use of, electric-ity with consequent effects on health, education, andproductivity.

To help address these issues, the focus of the gov-ernment’s energy policy is to promote activities thatwill increase access to electricity and provide a goodquality, cost-effective service while assuring the finan-cial viability of the economic agents engaged in pro-viding energy services and protecting the environment.To implement this strategy, the government plans to

32

3 The Energy Sector

The Energy Sector

33

carry out the following six key activities:1. Transfer the management and development of ELG

to a private operator and investor under a contrac-tual framework that provides incentives to improvethe operational and financial performance of ELG.

2. Revise the regulatory framework for the provisionof energy services in Rwanda.

3. Promote competition and support investment by theprivate sector to increase production of electricity.

4. Undertake to rehabilitate and expand infrastructureto meet national demand for electricity, keep coststo a minimum, and improve quality of service.

5. Encourage energy conservation through the rationaluse of energy and promote necessary measures toprotect the environment.

6. Promote rural electrification through both networkextension into rural areas and local power genera-tion. As a goal, the government’s 2020 Vision callsfor 36 percent of the population to be connected tothe grid by 2020.Gas sector policy is the responsibility of Unité de

Promotion et d’Exploitation du Gaz du Lac Kivu (De-partment of Promotion and Exploitation of MethaneGas from Lake Kivu, or UPEG), which used to be partof the former Ministry of Energy.UPEG reports to theMinistry of Infrastructure. UPEG is charged with de-veloping the methane resource by introducing PSPinto the sector, ensuring any such development is un-dertaken in a safe and sustainable fashion, and supervis-ing the establishment and use of gas-related assets, such

Figure 3.1 Energy: Existing and Future Key Roles and Responsibilities

Policymaking

Planning

Regulation

Ownership

Funding

Operation

Centralgovernment

Independentagencies Parastatals Corporate sector



Overall policy responsibilityrests with Ministry ofInfrastructure, Departmentof Energy

Economic regulation is currently the responsibilityof the government (Ministry of Infrastructure).Responsibility is to be transferred to the MultisectorRegulatory Agency once an Energy Law prescribingthe agency’s role with respect to the energy sectoris enacted.

Electricity assets and supply systems in urbanareas of Rwanda are currently owned by Electrogaz.The plan is for new generation assets to be ownedby the private sector. Partial or full sale of Electrogazto private sector is a possibility.

Effective responsibilityfor funding rests with thegovernment and aidagencies becauseElectrogaz operates on aloss-making basis.

Electrogaz currently operates and manageselectricity supply systems in most urban areas. Afive-year management contract is to be awardedto a private operator with the possibility of aconcession or outright sale following successfulcompletion. One or more methane gas extraction(and associated generation) concession contractsare likely to be agreed on in the near term.

Microgeneration and smallregional networks may beowned and operated at thecommunity level.

Rural electrification fundingwill require central govern-ment support but may beat least in part carried outby private organizationsand individuals.

Planning is presently the responsibility of the government (Ministry of Infrastructure). Depending on the natureof sector reform, responsibility may continue to be a central government function, may be transferred to bejointly held by the Multisector Regulatory Agency and a system operator or may be transferred entirely to theprivate sector.



34

as pipelines and storage infrastructure. There is somepreliminary consideration of turning UPEG into a pri-vate company, but no definitive plans have been an-nounced.

Electricity

GenerationTable 3.1 summarizes the nature, capacity, and outputperformance of Rwanda’s generation electricity port-folio, as well as its import and export transactions in2002. The key conclusions to be drawn from the tableare as follows:• Rwanda’s capacity portfolio is dominated by

hydroelectric plants, meaning that the country isvulnerable to both climatic and environmental fluc-tuations. A small diesel plant, Gatsata, which islocated in Kigali, is rarely used because of the pro-hibitive cost of imported fuel.

• A significant proportion (about 55 percent) ofRwanda’s electricity needs are met either throughimports or through Rwanda purchasing Burundi’sshare of the output of the Sinelac operation (ownedin equal shares by Burundi, the Democratic Re-public of Congo, and Rwanda).

• The total installed capacity for electricity genera-tion in Rwanda (28.55 megawatts) is not sufficientto meet peak demand, which is estimated to be40 megawatts.

• The output efficiency of the four hydroelectricplants is on average 38 percent. This low number

partly reflects the fact that the effective capacity ofthe plants is considerably below installed capacity. Itis low primarily because Ntaruka is used principallyat peak periods, owing to low lake levels (arisingfrom historic mismanagement of the reservoir) andbecause Mukungwa needs rehabilitation.Inspections by the independent consultants

Deutsche Energie-Consult Ingenieurgesellschaft mbh(DECON) in 1999 indicated that the generating plantwas in reasonable working condition, although a lackof preventive maintenance and unavailability of spareparts has meant that the plant has operated at less thanfull capacity.However, at the time this Country Frame-work Report was written, almost four years had passedsince DECON’s report was prepared, so the situationmay have deteriorated in the meantime.

Transmission and DistributionThe transmission network consists of some 285 kilo-meters of 110 kilovolt lines and 64 kilometers of70 kilovolt lines. A study conducted by Berocan in1998 indicated that the assets were in reasonable con-dition, requiring only some relatively minor conductorand insulator maintenance. The lack of spare trans-former capacity was, however, considered to make thenetwork vulnerable to outages.

The distribution system consists of both medium-voltage (30 kilovolt, 15 kilovolt, and 6.6 kilovolt) andlow-voltage (380 volt three-phase and 220 volt single-phase) networks, with a significant proportion beinglocated in Kigali and much of that sited underground.

Table 3.1 Electricity Provision in Rwanda

Installed capacity Output Percentage Output efficiencyPlant name Plant type (MW) (GWh) of total (percent)

Mukungwa Hydroelectric 12.5 56.691 25.8 45Ntaruka Hydroelectric 11.25 28.691 13.07 25Gihira Hydroelectric 1.6 6.912 3.15 64Gisenyi Hydroelectric 1.2 5.670 2.58 62

Subtotal (hydroelectric) 26.55 97.964 44.6 38Gatsata Diesel 2.0 0 0.0 0

Subtotal (all domestic sources) 28.55 97.964 44.6 36Imports (1.43) (0.7)Rusizi II (Sinelac) Hydroelectric 9 99.5 45.4Rusizi I (Snel) Hydroelectric 3.5 19.35 8.8Uganda Electricity Board 2.66 1.2

Subtotal 121.51 55.4Total 219.474 100.0


The Energy Sector

35

The Berocan study indicated that the distribution sys-tem was in a generally very poor state, with substationsin dangerous conditions, distribution lines lacking pro-tection from tripping, spare parts needed, and outdatedand incomplete information on the whereabouts of un-derground cables in Kigali that has resulted in frequentaccidental damage. All those factors produce a seriousdeleterious effect on supply quality and reliability.

System losses overall are estimated to be of theorder of 32 percent, with technical losses contributingsomething like 15 percent and commercial losses17 percent. This rate of system losses is very high byboth international and African standards.

SupplyAt its last official count, ELG had 43,177 electricitycustomers. Assuming this figure is an underestimate—which is a reasonable assumption, given the lack of in-formation ELG presently possesses concerning itscustomer base—the real number is somewhere in theregion of 50,000 customers. With an average house-hold size of six persons, this figure translates into a totalof just 300,000 persons receiving electricity—some3.8 percent of the national population.2 The MIEDuses a higher estimate than six persons per family totake account of large consumers of power, thereby giv-ing a higher figure for national coverage. Furthermore,these persons are principally urban residents;ELG is notresponsible for providing rural electricity service.

As a consequence, the vast majority of the energyneeds of Rwanda’s population are met through the useof other fuels, principal among them wood. This facthas important consequences both for the productivepotential and health of individuals and for the environ-mental condition of the country. The environmentalissue is addressed further below and in appendix A.

Little organized effort has been made to serve theelectricity needs of rural communities. There havebeen a number of studies on the use of off-grid elec-tricity solutions (including micro-hydro, solar, and bio-mass), and solar power is used by some hospitals andother bodies. ELG has also undertaken some limitedrural electrification projects, at the request and onbehalf of the government. The overall effect of theseprojects appears to have been negligible.

A single flat rate tariff of FRw 42 (approximatelyUS$0.082) per kilowatt-hour was set by the Ministryof Infrastructure in 1998 and has not changed since.

This tariff does not reflect actual costs. A single tariffstructure means both that there is no “life-line tariff ”for the poorest customers and that there is little incen-tive to encourage energy conservation (such as mightbe achieved by means of a rising block tariff structure).

Obtaining comprehensive, up-to-date, and reliabledata is relatively difficult. Nonetheless, figure 3.2 pres-ents some basic benchmarking data concerning varioustechnical and service parameters. The data indicatethat Rwanda falls short of many of its African neigh-bors in terms of level and quality of service.

Financial and Human Resource PerformanceChapter 5 comments on the financial performance ofELG and its weaknesses with respect to billing andcollection. The chapter also contains a summary ofELG’s management and human resource capability, aswell as some comments on the structure and contentof the management contract that is currently undernegotiation.

The Gas Sector

Operational assets in Rwanda’s gas sector are currentlylimited to a single, small-scale, methane gas extractionfacility on Lake Kivu. This plant was constructed in1963 as an experimental facility but has for some con-siderable time constituted no more than an energysource to fire the boilers of a local brewery.

A number of feasibility studies have been under-taken with respect to large-scale extraction of methanegas from the lake and its subsequent use for power gen-eration purposes.An estimated 50 billion to 56 billioncubic meters of methane are thought to be available forexploitation—a resource of enormous potential signif-icance for Rwanda. At the time we prepared this re-port, two consortia had expressed interest in projectdevelopment, and one was in the early stage of conces-sion negotiations with the government. One proposalentails constructing a single methane extraction facilitywith an associated 25 to 30 megawatt generation plant.The other envisages multiple extraction facilities alongthe lake, with associated 8 to 10 megawatt generators.

The lake’s potential for power generation wouldbenefit mainly urban residents who are hooked up tothe ELG distribution system.The exploitation of LakeKivu gas, however, also offers the prospect of provid-ing liquefied natural gas supplies to the wider rural


36

Figure 3.2 Benchmarking Data in Electricity

C. System losses D. Access to electricity

05

10152025303540

Percent

Number US ¢ per kilowatt hour

A. Generation capacity

B. Electricity consumption per capita

Angola

Botsw

ana

Buru

ndi

Congo

, Dem

.

Rep. o

fKen

ya

Madag

ascar

Malawi

Maurit

ius

Mozam

bique

Rwanda

Seyc

helle

s

Swaz

iland

Tanza

nia

Ugand

a

Zambia

Zimba

bwe

0

500

1,000

1,500

2,000

2,500

3,000

Megawatt

Kilowatt hour

Botsw

ana

Botsw

ana

Buru

ndi

Angola

Comor

os

Congo

, Dem

.

Rep. o

fKen

ya

Madag

ascar

Leso

tho

Maurit

ius

Malawi

Mozam

bique

Namibi

a

Rwanda

Seyc

helle

s

Swaz

iland

Tanza

nia

Ugand

a

Zambia

Zimba

bwe

0

600

400

200

800

1,000

1,200

1,400

1,600

Buru

ndi

Kenya

Malawi

Rwanda

Tanza

nia

Ugand

a

Zimba

bwe

010203040506070

Population (percent)

Kenya

Malawi

Rwanda

Sout

h Afri

ca

Zimba

bwe

E. Customers per employee F. Average tariff

020406080

100120

Kenya

Malawi

Rwanda

Sout

h Afri

ca

Zimba

bwe

0

2

4

6

8

10

Kenya

Malawi

Rwanda

Sout

h Afri

ca


The Energy Sector

37

population for whom, if economical, its use would rep-resent a much healthier and more environmentallyfriendly alternative to burning wood.

Investment Needs and Priorities

Several needs and priorities have been identified withrespect to the energy sector.

RehabilitationThe DECON study provides a detailed and compre-hensive, although now somewhat outdated, assessmentof the state of each of the generating plants owned andoperated by ELG. It contains a number of proposals asto where the replacement, upgrading, and mainte-nance of specific aspects of the plant would increaseoperating efficiency and reduce outages. Table 3.2contains summary estimates of base rehabilitationcosts for three generators. Berocan undertook similarwork to assess and cost the rehabilitation and mainte-nance requirements for transmission and distributionnetworks in five urban areas. As is evident fromtable 3.2, the majority of the rehabilitation funding istargeted at the networks and the low-voltage net-works in particular.

New Generation Investment No new generating plant hasbeen constructed in Rwanda since 1982, and given thecurrent supply shortage, new investment in electricitygeneration is evidently needed. A number of proposalshave been put forward and are under considerationby the government, as summarized in table 3.3. LakeKivu is regarded as the best alternative for a number ofreasons, including its relatively low construction cost

and short lead time, its relatively low estimated operat-ing costs (estimates of US$0.05 per kilowatt-hour com-pare favorably with hydroelectric power), its potentialfor serial development in response to Rwanda’s chang-ing needs, its limited environmental effect, and finallyits diversification away from hydroelectric power.

New Network Investment A number of proposals havebeen put forward to extend the high- and low-voltagenetworks.The priority, however, should be to connectany new generation facility (a 20 to 30 kilometer linewould be required to connect a single Lake Kivu sta-tion to the grid) and to connect as many large indus-trial customers—such as the tea estates—as possible tohelp reduce wood fuel consumption. It should benoted that Rwanda’s challenging topography meansthat network construction costs are relatively high andcan reach up to US$80,000 per kilometer.

Table 3.2 Rehabilitation of Assets in the Energy Sector

Asset requiring rehabilitation Total cost (US$)

Mukungwa hydroelectric generating station 1,950,000Gihira hydroelectric generating station 1,465,000Gisenyi hydroelectric generating station 1,477,000

Generation total 4,892,000High-voltage lines and substations 13,214,000Medium-voltage lines (over and underground) 5,698,000

and substationsReinforcement of medium- and high-voltage 5,300,000

substations and switchboxesLow-voltage lines and substations 13,829,000

Transmission and distribution total 38,041,000Overall total 42,933,000

Note: Total cost includes various types of costs, including direct and indirectcosts as well as costs for improvements, materials, and installation, dependingon the type of expenditure involved.


Table 3.3 Proposals for New Generation Investment

Rwanda Lead Approximate CapacityRelative share Capacity time project cost cost

Plant priority (percent) Fuel (MW) (years) (US$) (US$/MW)

Lake Kivu 1 100.0 Gas 25.0 2 25,000,000 1,000,000Rusomo Falls 2 33.3 Hydroelectric 61.5 5 170,000,000 1,500,000Nyabarongo 3 100.0 Hydroelectric 28.0 4 77,000,000 1,500,000Rusizi III 4 33.3 Hydroelectric 82.0 5 170,000,000 2,100,000Source: Adam Smith Institute research.


38

Potential Sources of Investment Financing

Several sources of investment financing have been iden-tified, or secured, to pay for the various rehabilitationwork and new capital works discussed above, as summa-rized in table 3.4. With respect to urban electricity sup-ply, the World Bank’s short-term loan is already in place,while the loan from the African Development Bank,the loan from the Arab Bank for Economic Develop-ment in Africa and Organization of Petroleum Export-ing Countries, and longer-term World Bank loans haveall been approved and now await the investment plan tobe prepared by ELG’s future management contractor.Consequently, considerable funding appears to be avail-able. Even if it is assumed that rehabilitation cost esti-mates will need to be increased significantly because ofthe time that has elapsed since their preparation, theamount of donor funds available appears sufficient. Interms of private investment for the Lake Kivu genera-tion project, the existence of two consortia interested inthe project also bodes well.

Opportunities for Private Sector Participation

There appear to be five main opportunities for engag-ing the private sector in owning, operating, and en-hancing the use of assets in Rwanda:1. The five-year management contract for running

ELG is clearly the most immediately significant andimportant form of PSP affecting the energy sector.In principle, after the contractor has improvedELG’s operational and financial performance, it

should be possible to establish deeper forms of PSPinvolving a significant funding commitment by theprivate sector, most probably in the form of a long-term concession contract, although an outright saleis also possible. It is not yet clear whether such atransaction might best involve the whole of ELG orseparate electricity and water functions.

2. The serious need for additional generation capacityrepresents the next most immediate priority forPSP. The granting of a concession for a methanegas extraction plant and an associated generationfacility represents the best, and most likely, prospectfor introducing a private sector operator into theindustry.

3. PSP will certainly be required if a far-reaching ruralelectrification program is going to be implemented.In particular, the development of microgenerationand regional networks is most likely to require pri-vate sector initiatives, with financial support fromthe government or donors. National grid extensioncould also involve PSP.

4. The potential exists to contract certain service ele-ments out to the private sector. Clearly, proper pro-curement rules will need to be put in place first, butsuch a program would have the added benefit ofencouraging the development of local skills andcapability.

5. In principle, some form of private sector involve-ment in running the emerging MSR might also bepossible.The MSR faces an uphill task in terms ofthe volume of work it already faces while simulta-neously trying to build capacity.

Table 3.4 Sources of Investment Financing

Source of financing Amount (US$) Comment

African Development Bank 29,000,000 US$15 million of this loan is for rehabilitating Kigali’sdistribution network and national transmission lines.

Arab Bank for Economic Development 10,500,000 Loan is for rehabilitating three hydroelectric plants.in Africa and Organization ofPetroleum Exporting Countries

World Bank 7,770,000 Loan consists of an ELG management contract and ashort-term investment to improve operational andcommercial capability of ELG.

World Bank 30,000,000–40,000,000 Loan represents a longer-term investment for rehabilitationof water and electricity assets.

Private Not yet confirmed This private sector investment is for the construction of amethane gas–fired independent power plant at Lake Kivu.


The Energy Sector

39

Issues

Several issues must be addressed.

Success of ELG Management ContractIt is understood that agreement has been reached onthe terms of the ELG management contract, althoughthe contractor is not yet in place. Completion of thisstep represents a clearly positive development for theenergy sector.The particular terms of the contract re-mained private between the parties at the time thisreport was drafted, however, and it is not therefore ap-propriate to comment on them here. In general terms,though, it is clear that, if the long-term objectives ofthe government are to be realized, the performanceregime must provide a framework of incentives thatstrongly encourage the management contractor to im-prove revenues and to minimize costs while achievingsatisfactory standards of service. A remuneration for-mula that relies as far as reasonably possible on variablepayments linked to, among other things, collectionrates, technical and commercial losses, and service lev-els will have significant advantages over an approachthat incorporates a high proportion of fixed payments.For these incentives to be effective in practice, a strongand equitable mechanism for monitoring and compli-ance must also be in place.

It will also be important to make adequate invest-ment funding available to the management contractor.In this respect, it will be most important that the majorloan proposed by the World Bank for ELG asset reha-bilitation be regarded as an absolutely firm and reliablecommitment.

Other loans tied to asset rehabilitation identifiedabove are, reportedly, secure.Pending the completion ofnegotiations, it is not yet fully clear what the manage-ment contractor’s role will be in terms of utilizing thesefunds.The provision of funds is to some extent contin-gent on investment plans and proposals to be preparedby the management contractor during the initialcontract period. The uncertainty that flows from thisapproach is unavoidable, given the impossibility of es-tablishing a reliable assessment of investment needs andpriorities before the management contractor is in place.The importance of ensuring that the managementcontractor is subject to strong and clear performanceincentives cannot be overemphasized. However, this

aspect of the contractual framework highlights the factthat the arrangements proposed over the next five yearsshould also be seen as a collaborative partnership be-tween the private sector, government, and donors.

Legal and Policy FrameworkIt is imperative that an energy law (or separate elec-tricity and gas laws) be prepared as soon as possible.3

The law (or laws) will have two principal purposes:first, to set out the role and responsibility of the MSRwith respect to the energy sector, and second, to setout a policy framework for the sector.This frameworkshould, at a minimum, incorporate the following keyelements:• A statement of overall government policy for the

sector—that is, principles, objectives, and targetsfor sector development covering at least the next10 years

• A roadmap setting out important milestones for in-dustry development

• A vision for the structural and operational evolu-tion of the electricity sector, including a clear anddetailed description of the wholesale tradingarrangements envisioned for the industry; of ELGand UPEG’s structure, role, and ownership in thisnew environment; of the position of the privatesector; and of how competition will be managed

• A vision for the structural and operational evolu-tion of the gas sector and how its close interrela-tionship with the power sector will be managed

• Details concerning the nature and application ofthe government’s subsidy policy.This policy framework is necessary for the govern-

ment and ELG to understand how the energy sectorwill evolve over time and what steps need to be takenand when the vision will be achieved. It is also impor-tant for private sector investors, who will wish to un-derstand the potential for and nature of their prospectiveparticipation in the industry.Without such a framework,contract negotiations with any private investor (such asa potential methane gas concessionaire) will prove con-siderably more difficult: the government will not knowwhat sort of flexibility will need to be built into thecontract to accommodate sector evolution, and theprivate operator will want financial compensation forany additional perceived risk incurred from a lack ofknowledge concerning industry development.


40

Multisector Regulatory AgencyIt is also important that the function, responsibility, andjurisdiction of the MSR—as far as energy sector oper-ation is concerned—be specified as soon as possible.For example, the following questions should be clearlyaddressed:• Should the Ministry of Infrastructure, UPEG, or

MSR be responsible for determining rules for ac-cess to, and pricing of, gas pipeline and storage fa-cilities as and when they are constructed?

• How many licenses should ELG have? The introduction of a regulatory framework will not

in itself reassure potential investors. Rather, it is thecreation of a truly independent, properly resourced, andeffective regulatory body that investors will wish to see.The present constitution of the MSR leaves room fordoubt concerning some of these important questions.For example,the MSR Board,which holds executive re-sponsibility for delivering regulatory policy,is composedentirely of civil servants,which means the MSR appearsnot to be properly independent from the government.In addition, the MSR itself has not been granted appro-priate powers—for example, in the telecommunicationssector the MSR is not charged with licensing all com-panies. In the absence of an appropriate regulatoryframework, investors may still appear, but their engage-ment will likely require a very inflexible contract or apotentially unaffordable financial premium.

Lake Kivu GasThe successful conclusion of a concession agreementwith one of the consortium seeking to develop themethane gas resource for power generation purposes isclearly a short-term priority given Rwanda’s shortageof electricity supply.Two separate agreements with thetwo consortia are unlikely because there is insufficientdemand in the short term for the gas extracted.Consid-erable care must be taken in executing this agreementbecause—as a long-term contract—it will affect boththe potential for success in awarding a concession agree-ment for ELG in five years and the implementation ofnew wholesale trading arrangements for the industry.

Rural ElectrificationThe Ministry of Infrastructure needs to develop a ruralelectrification plan as soon as is feasible. This planshould complement the energy law described above

but should specify in considerable detail how energyservices will be extended to rural areas, includingmechanisms for government support and subsidies tothe private sector to enable the development of micro-generation and network development initiatives. Suchfinancial support may be delivered by means of• The creation of a rural fund• Direct payments to operators• Indirect assistance through tax or import duty

incentives.The plan should make use of the studies and initia-

tives that have been performed to date, such as theMukurange electrification project, which, although notentirely successful, does provide important lessons interms of what constitutes a commercially viablescheme. Further policy studies are also needed to aidunderstanding of what customers are willing and able topay and to pinpoint development opportunities.

There may also be a case for promoting greatercommunity and nongovernmental organization in-volvement in the development of rural electrificationschemes. Such an approach would be likely to achievethe following:• Better value for the potentially available subsidy

because closer community involvement wouldencourage provision at an appropriate standard tothose who will benefit most

• A higher level of long-term sustainability.The essential features of a possible scheme focused onrural water service are set out in chapter 4.

Recommendations

The preceding discussion leads to the following rec-ommendations for facilitating and making best use ofPSP in the energy sector:• Prepare an energy policy framework that clearly

sets out how the sector will evolve with respect toits structure, ownership, and trading arrangements.This framework should set out unambiguousmilestones for achieving this vision and should beused to assist with independent power producernegotiations.

• Draft and enact new electricity and gas laws—or acombined energy law—that incorporates theenergy policy framework and sets out the roleand responsibility of the MSR with regard to

The Energy Sector

41

supervising energy sector operations.The regulatorshould be granted sufficient power and resources tomake it as independent from government as possi-ble while still being required to conduct its activi-ties in a transparent manner. Revision of the MSRlaw may also be necessary to ensure either that theexecutive responsibility for delivering regulatorypolicy rests with those actually creating policy orthat the constitution of the board is altered so thatit is no longer dominated by representatives fromgovernment. Private sector operators need to beconvinced that the regulator will act impartially toprotect their interests.

• Proceed with the methane gas concession agree-ment but preferably within the context of the en-ergy policy framework and, if possible, following

Notes

1. The majority of the information here isderived from a report prepared for theRwanda Investment Promotion Agency bythe Africa Institute for Policy Analysis andEconomic Integration (AIPA 2002) titled“The Rwandan Economy: A Strategy forInvestment.” The RIPA report summarizesmuch of the key content of Booz-Allen &Hamilton’s 1999 “Rwanda Diagnostic Re-port” as well as a 1998 report by DeutscheEnergie-Consult Ingenieurgesellschaft mbh(DECON) reviewing generation rehabilita-

tion costs and Berocan’s 1998 report review-ing network rehabilitation costs. Each ofthese documents may be referred to if fur-ther information is required.2. The estimated penetration of electricityas a percentage of the total population de-pends on an estimated household size of six.Clearly, a higher percentage penetration willresult as average household size increasesor significant institutions are taken intoaccount. The figure may, in a best-casescenario, be as high as 6 percent of the

population receiving electricity, given thisvariation.3. A draft gas law has been prepared, but it isnot a law as such because it contains very lit-tle policy for the gas sector. Rather it is adraft contract for a service agreement, andthis draft itself needs amendment in light ofplans to award a concession contract, ratherthan a service contract, for Lake Kivumethane gas development.

the passage of the gas and electricity (or combinedenergy) laws.

• Prepare a rural energy plan that sets out a strategyfor implementing rural electrification, completewith measurable objectives, a clear funding policy,and a prioritized action plan that the private sectorcan potentially engage with.

• Carry out a feasibility study focused on the possi-bility of implementing a demand-led scheme forfunding rural electrification.Although sufficient donor funding appears to be

in place for asset rehabilitation and maintenance, itis important that effective donor coordination takesplace to ensure that other important elements underly-ing sector development, including the rural electrifica-tion and environmental policy initiatives, proceed.

As recognized both in the United Nations MillenniumDeclaration and at the 2002 Johannesburg summit, theprovision of adequate water supply and sanitation serv-ices has a more direct effect on human health and welfarethan the provision of any other infrastructure service andcontributes directly to the well-being of the poorest insociety. Improved access to safe water supplies not onlygenerates immediate benefits in terms of significantmeasurable reductions in morbidity and mortality rates,but also frees up a substantial proportion of the time of(especially) women and children for more productiveuse. Improved sanitation also brings immediate benefitsin terms of both public health and the environment.

Despite their particularly direct effect on povertyalleviation, improvements in the provision of water andsanitation services generally have a more diffuse effecton economic growth and income generation than doimprovements in the other infrastructure services con-sidered in this report.

This chapter first considers the institutional frame-work for water and sanitation services in Rwanda. Itthen assesses the current performance of the sector inboth urban and rural areas and recommends measuresto improve this performance—in particular throughthe promotion of private sector participation (PSP).


Overall responsibility for water resource managementpolicy in Rwanda resides in the Ministry of Infrastruc-ture. The allocation of key roles and responsibilitiesrelating to the urban and rural water supply sectors as

they are currently organized and managed in Rwandais summarized in figures 4.1 and 4.2, respectively.

Water supply and sanitation policy in Rwanda isthe general responsibility of the Department of Waterand Sanitation in the Ministry of Infrastructure. InKigali and 13 other urban areas, responsibility forthe provision of potable water supplies resides withElectrogaz (ELG). The districts hold responsibility forwater supply in rural areas, although this responsibilityis generally delegated to a community level in practice.Urban areas served by ELG are as follows: Butare,Byumba, Cyangugu, Gikongoro, Gisenyi, Gitarama,Kibungo, Kibuye, Kigali, Nyagatare, Nyanza, Ruhango,Ruhengeri, and Rwamagana.

Responsibility for sanitation services resides at thedistrict level in both urban and rural areas, although inpractice almost no service is actually provided.


The core policy objectives for the water supply andsanitation sector are• To improve the provision of water and to extend

the water supply network• To increase access to sanitation services• To encourage community participation in the in-

stallation and management of water and sanitationinfrastructure

• To promote technically and financially viable proj-ects based on strong community participation

• To strengthen capacity at both the central govern-ment and the district level.

42

4 The Water and Sanitation Sector

The Water and Sanitation Sector

43

These policy objectives are not currently set out inany formal water and sanitation policy statement,although a policy document has been under prepara-tion for some time, and a second draft was beingconsidered by the government at the time this reportwas prepared.

Urban Water SupplyWithin the city of Kigali and the other main urbancenters in Rwanda, the principal means that the gov-ernment is using to achieve the above objectives in thepotable water subsector is the introduction of PSP. Thiseffort will initially be achieved by the award of a five-year management contract for ELG. The process ofdesigning and implementing this contract has been

substantially completed. At the time this report wasdrafted, an agreement was reported to have beenreached with a preferred bidder, although the contrac-tor is not yet in place.

The PSP scheme about to be put in place does notentail any private sector contribution to funding theimprovement and extension of ELG’s water supplyfacilities.The intent is that over the period of the con-tract ELG’s commercial and operational performanceshould improve to the point where deeper forms ofPSP will become feasible in the longer term, therebyopening up the possibility that substantial private sec-tor financing toward improved service provision andnetwork extension might be obtained under a long-term concession, to be introduced subsequently.

Figure 4.1 Urban Water Supply: Roles and Responsibilities

Policymaking

Planning

Regulation

Ownership

Funding

Operation

Centralgovernment




Overall policy responsibilityrests with the Ministry ofInfrastructure, Departmentof Water and Sanitation.

Economic regulation is currently the responsibilityof the government (Ministry of Infrastructure).Responsibility is to be transferred to the MultisectorRegulatory Agency once it has established sufficientcapacity to accept the role.

Water supply systems inurban areas of Rwandaare currently owned byElectrogaz.

Effective responsibility forfunding rests with thegovernment and aidagencies becauseElectrogaz operates ona loss-making basis.

Electrogaz currently operates and manages watersupply systems in most urban areas. A five-yearmanagement contract is to be awarded to aprivate operator.

Sanitation provision is leftto private organizationsand individuals. There areno reticulated seweragesystems currently inoperation.

Sanitation funding is largelya matter for privateorganizations andindividuals.

Electrogaz is currently responsible for planning. Thisresponsibility will be picked up by the new managementcontractor.



44

Funding for the rehabilitation of existing facilitiesduring (approximately) the first two years of the man-agement contract and funding for further developmentof the water supply system (projects identified in in-vestment plans prepared by the management contrac-tor) during the remaining years of the contract willtherefore be externally provided using donor funds.Nevertheless, even within the term of the managementcontract, the approach that is being adopted allows forthe possibility of raising additional private sectorfinancing to support individual development projectsunder build, operate, and transfer (BOT) or similararrangements.

Apart from the Kigali effort, a project is planned forthe reinforcement of Butare’s water supply system, and

the government has requested donor funding for stud-ies of the needs of other urban centers.

Urban SanitationIn the sanitation subsector, specific policy measuresfocused on the urban areas are not currently in evi-dence. The Atelier Technique sur la Politique Secto-rielle de l’Eau et de l’Assainissement (Workshop onWater Supply and Sewerage Sector Policy) held inSeptember 1997 agreed on the following strategy:• Establish an appropriate institutional and legal

framework and finalize and formalize the Sanita-tion Code.

• Adopt the “polluter pays” principle.• Protect natural resources and the environment.

Figure 4.2 Rural Water Supply: Roles and Responsibilities

Policymaking

Planning

Regulation

Ownership

Funding

Operation

Centralgovernment

Districtauthorities Parastatals

Small businesssector



Overall policy responsibilityrests with the Ministry ofInfrastructure, Departmentof Water and Sanitation.

Water supply systems inrural areas are owned bythe districts.

Funding is primarily bycentral government usingdonor funds but also byNGOs using bilateral aid.

Under the Water andSanitation Program, thecommunity funds 10% ofcapital costs. NGO-ledschemes typically involvesimilar community funding.

Responsibility for schemeplanning is held at a districtlevel, but the Departmentof Water and Sanitationassists with complexschemes

Charges for simple schemesare set by communitycommittees.

Responsibility for themajority of schemes isdelegated to communitycommittees, which appointsomeone to look after thefacilities.

Responsibility for schemedesign is delegated to thecommunity level wherepossible.

Charges for complexschemes are set bythe districts.

District authorities areformally responsible foroperation of all schemesbut operation of simplerschemes is usually delegatedto a community level.

There is provision foroutsourcing maintenanceresponsibility to privateengineering companies.



45

The following action plans were also agreed on:• Publish and disseminate the Sanitation Code.• Work out a strategy for the most densely populated

urban areas.• Require the preparation of a collective or shared san-

itation scheme for all new areas under construction.• Promote the harmonization of the processing and

management of waste by the most economic over-all means available.

• Put in place or improve stormwater drainage systems.• Develop a plan for removing and treating industrial

wastewater.• Carry out a pilot project to test the applicability,

in the Rwandan context, of the demand-led sanita-tion strategies adopted in Kumasi, Ghana, andOuagadougou, Burkina Faso, with the objective ofadopting and applying a similar approach based onwillingness to pay.

• Identify all participants in the subsector and definetheir roles, following the example of the nationalpolicy on housing.

• As soon as possible, run education, information, andsensitization campaigns on hygiene.

• Manage solid waste in a fashion that permits recy-cling and processing into useful products.

Rural Water Supply and SanitationDecentralization lies at the heart of the government’sapproach for water supply in rural areas.The objectiveof decentralization is to promote the development ofdemand-led schemes initiated at the community level.It is focused on ensuring that water infrastructure is de-veloped in ways that meet the community’s prioritiesat charges that are at the same time both affordableand sufficient to cover long-term operating andmaintenance costs.A mobilization process designed toassist communities in establishing community waterassociations—and aid them in agreeing on their re-quirements and in planning—forms an essential earlystage of the government’s approach.

Under the decentralization model,• Policy formulation is the responsibility of the gov-

ernment, which also provides the majority of initialfunding for capital expenditure.

• Planning and maintenance and operations are theresponsibility of the community, which must alsomake a contribution to initial capital costs.

• System ownership resides at the district level.

A number of specific initiatives have been estab-lished to improve water and sanitation infrastructure inrural areas:• The Rural Potable Water and Sanitation Project,

which entails a six-year program to support thedevelopment of water and sanitation services in11 districts.The project is financed by a US$20 mil-lion International Development Association credittogether with an intended US$1 million commu-nity contribution and US$500,000 of governmentfinancing. Three pilot projects are currently inprogress.

• A project funded by Kreditanstalt für Wiederaufbau(Kf W) to establish potable water supplies and sani-tation services in eight districts in Kigali Rurale.

• The water and sanitation component (US$10 mil-lion) of the Umutara Community InfrastructureDevelopment Project, which is directed toward theprovision of services in all seven districts in theprovince (currently at the planning stage).

• Government-funded schemes for the provision ofpotable water in the districts of Mukingi andBwisige (at a cost of FRw 214 million andFRw 152 million, respectively).

• The Murambi Potable Water Supply Project, whichcovers another Umutara district and is due forcompletion early in 2003. The project is receiv-ing Japanese and German (KfW) funding ofUS$589,000 and US$300,000, respectively.

• A scheme for the possible integration of theKarenge water supply system with that of Ngendato cover the entire Bugasera region.The project isfunded by the European Union at an estimated cost€19.3 million.The Ngenda water supply system iscurrently operated by a private company (SHERIngénieurs-Conseils) under a form of BOT buthas excess capacity of treated water from LakeCyohoha. Bid documents are being prepared andfinancing has been agreed on in principle.

Sector Performance

Overall access to potable water in Rwanda is low byAfrican standards,with some 52 percent of the nationalpopulation estimated as having access to safe water(defined in terms of the percentage of the populationliving within 500 meters of a safe water source).


46

Sanitation services are largely limited to individual on-site provision and, taken as a whole, are poor.

Urban Water SupplyAccording to available statistical data, the performanceof ELG in the urban water sector appears to be weakby comparison with that of a selection of other Africanwater utilities.1 ELG has one of the highest levels ofunaccounted-for water in Africa, although it is notclear to what extent this statistic arises as a result oftechnical losses as opposed to administrative losses (thatis, failure to meter or bill consumption because of

illegal connections and poor information and com-mercial systems). ELG also achieves one of the lowestratios of customers per employee among African utili-ties (although it should be noted that a high proportionof ELG connections take the form of shared standpipeconnections). Figures 4.3 and 4.4 show the relevantnumbers.

ELG’s current tariff structure is shown in table 4.1.Establishing a reference value for tariff comparisonswith other African utilities is made difficult by the factthat the Rwandan franc has depreciated substantiallyagainst the U.S. dollar in recent years. The dollar

Figure 4.3 Unaccounted-for Water in African Water Utilities

0

10

20

30

40

50

60

70Percent

AWB

LNW

SODEC

ICoT

DMWS

WUC

UWSA

WASA PA SD

E

MANGAUNGDW

D

TUW

ASA

DAWASA

GCC

TUW

SA

MWSC

BWB

AUWSA

CWSC

NWCPC

NWSC

DUWASA

SHUW

ASA

IRUW

ASA

NYEWASC

O

NAQW

ASS BWKW

SC ELG

LuW

SCGW

CL

NCC-WSD

AHC-MMS

MWSA

Source: Water Utility Partnership SBNet 2001.

Figure 4.4 Number of Connections of African Water Utilities

0

50

100

150

200

250

300

350Customers per employee

SODEC

IEL

G

NWSC

SHUW

ASA

IRUW

ASA

DUWASA

WUC

BWB

GWCL

MWSA

UWSA

WASA

CWSC

NAQW

ASS

AHC-MMS

AUWSA

DAWASA

TUW

ASA

KWSC

NYEWASC

O

LuW

SC

TUW

SA

MWSC

NWCPC BW PA

SDE

MANGAUNGGCC

DMWS

Source: Water Utility Partnership SBNet 2001.


47

equivalent rate per cubic meter has slipped from 0.63to 0.37 for the lowest consumption band since the tar-iff was introduced in 1997. At US$0.63 per cubicmeter ELG’s first block price is extremely high byAfrican standards; at US$0.37 per cubic meter, theprice is more in line with African norms. Overall aver-age revenue for ELG water in 2001 was FRw 285 percubic meter, or US$0.55–0.89 (at the current and 1997exchange rates, respectively).A comparison with otherAfrican water utilities is given in figure 4.5.

Table 4.1 Electrogaz Water Tariffs

1997–present

Monthly consumption FRW/m3 USD/m3 USD/m3

1997 20030–25 m3 200 0.63 0.3726–60 m3 300 0.94 0.5661–100 m3 350 1.09 0.65Over 100 m3 375 1.17 0.70Standpipe concessions 200 0.63 0.37Source: Electrogaz.

Figure 4.5 Average Price of Water Supplied by African Water Utilities

0

0.20

0.40

0.60

0.80

1.00

1.20

1.40Average revenue (US$/m3)

WUC

SDE

ELGW

ASABW

B

NYEWASC

O

NWSC

MANGAUNGBW

NWCPC

IRUW

ASA

TUWSA

DAWASA PA

MWSA

AWB

CWSC CoT

NCC-WSD

TUWASA

GWCL

AUWSA

SHUW

ASA

NAQW

ASS

AHC-MMS

GCC

LuW

SC

DUWASA

DWD

KWSC

UWSA

Typical U.K. equivalent

Note: The top graph shows ELG’s average revenue (midpoint value) in U.S. dollars compared with a selection of other African water utilities.The bottom graph againshows comparative average revenue and also provides an indication of how the total cost of typical per capita water consumption compares with average grossnational product (GNP) per capita in the country in which each utility is based. This comparison is based on a nominal per capita consumption of 40 liters per day(equivalent to approximately 15 cubic meters per year). This shows particularly clearly how expensive ELG water supply is in relation to per capita income levels inRwanda.Source: Water Utility Partnership SBNet 2001, supported by GNP data obtained from the CIA World Factbook.

0

0.20

0.40

0.60

0.80

1.00

1.20

1.40

Average revenue (US$/m3)Cost of minimal annual consumption

(percent of per capita GNP)

5.0

4.5

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0

ELG

BWB

SDE

NWSC

NYEWASC

OW

ASA

IRUW

ASA

TUWSA

DAWASA

MWSA

NWCPC

TUWASA

CWSC

NCC-WSD

AUWSA

SHUW

ASA

GWCL

NAQW

ASS

AHC-MMS

WUC

LuW

SC

DUWASA

DWD

GCC

KWSC

MANGAUNGBW PA

AWB

CoTUW

SA

Average revenue Cost per year compared with per capita GNP

Rwanda: Cost per year compared withGNP per capita

Rwanda: Average revenue (US$/m3)


48

In practice, the majority of the urban populationpays substantially more for water than the ELG averagerate.Water is currently sold to standpipe operators at aflat rate of FRw 200 per cubic meter and is then resoldto final users.There is no control over the prices thatstandpipe operators charge, but the normal pricecharged in Kigali was about FRw 10 for a 20-literjerry can at the time of preparing this report—or FRw500 per cubic meter (that is, US$0.97–1.56).This priceis exceptionally high. It is the poorest in society whoare the most exposed to these high prices, because theyhave no safe alternative but to use standpipe water,paying almost three times the amount that wealthierhouseholds pay for water to meet essential needs.Figure 4.5 shows average revenue per cubic meter ofwater sold by each utility and does not take into ac-count the supplementary amount paid to standpipeoperators. It clearly demonstrates the exceptionallyhigh price of ELG supplies.

Despite high tariff levels, ELG’s financial perform-ance remains extremely weak, although it is impossibleto ascertain to what extent this weakness is attributableto water supply as opposed to electricity supply activi-ties. In part, it reflects the company’s weak financial sys-tems, especially in relation to customer management.

It is estimated that less than 70 percent of water servicebilled is collected, and unbilled consumption is likelyto represent a significant proportion of unaccounted-for water. Inefficiency is also evident in operations,particularly in relation to overemployment.

In addition to the need to improve its overall oper-ational and commercial performance—issues to be ad-dressed within the framework of the managementcontract that is currently under negotiation—ELGfaces a number of fundamental difficulties that willneed to be addressed in the near future. Principalamong these is the poor quality of the main watersource for Kigali, the Yanze River.The extremely highlevels of sediment carried by the river, especially attimes of high flow, result in significant operationalproblems in filtration and treatment. During periodsof low flow, ELG must take almost the entire flow tosatisfy demand from the city and there is a significantoverall supply deficit. Also, the five main springs onwhich Kigali also relies for a significant proportion ofthe balance of its water supply, especially during thedry season, are centrally located and have become en-veloped as the city has expanded.They are accordinglyhighly vulnerable to pollution. Finally, growth in pop-ulation—and hence in demand—has also put serious

AHC-MMS AHC Mining Municipal Services Ltd. Zambia

AUWSA Arusha Urban Water Supply andSewerage Authority Tanzania

AWB Amatola Water Board South Africa

BW Bloem Water South Africa

BWB Blantyre Water Board Malawi

CoT City of Tygerberg South Africa

CWSC Chipata Water and Sewerage Company Zambia

DAWASA Dar es Salaam Water and SewerageAuthority Tanzania

DMWS Durban Metro Water Services South Africa

DUWASA Dodoma Water and Sewerage Company Tanzania

DWD Department of Water Development Tanzania

ELG Electrogaz Rwanda

GCC City of Gweru Zimbabwe

GWCL Ghana Water Company Ltd. Ghana

IRUWASA Iringa Urban Water Supply and Sewerage Tanzania

KWSC Kafubu Water and Sewerage Co. Ltd. Zambia

LNW Lepelle Northern Water South Africa

LuWSC Lusaka Water and Sewerage Co. Zambia

MANGAUNG Mangaung Local Municipality South Africa

MWSA Mwanza Water and Sewerage Authority Tanzania

MWSC Mulonga Water and Sewerage Company Zambia

NAQWASS Nakuru Water Company Kenya

NCC-WSD Nairobi City Council Kenya

NWCPC National Water Conservation and Pipeline Co. Kenya

NWSC National Water and Sewerage Company Uganda

NYEWASCO Nyeri Water and Sewerage Co. Ltd. Kenya

PA Drakenstein Municipality South Africa

SDE Senegalaise des Eaux Senegal

SHUWASA Shinyanga Urban Water and SewerageAuthority Tanzania

SODECI Societe de Distribution d’Eaux de Côte d’Ivoire Côte d’Ivoire

TUWASA Tabora Urban Water and SewerageAuthority Tanzania

TUWSA Tanga Urban Water Supply andSewerage Authority Tanzania

UW Umgeni Water South Africa

UWSA Urban Water Supply and SewerageAuthority Tanzania

WASA Water and Sewerage Authority Lesotho

WUC Water Utilities Corporation Botswana

Key to Figures 4.4 and 4.5.


49

pressure on the overall adequacy of the available waterresources, although this problem may in part be due tothe high level of estimated leakage from the distribu-tion system.

The poor quality of existing resources for Kigaliand the increasing supply deficit mean that there is astrong case for developing alternative sources of supply.The most promising potential new resource appears tobe water piped under gravity from catchments in theareas of Rubindi, Mutobo, and Mpenge in the Virungaregion.The project would involve a pipeline of at least800 millimeters in diameter, running some 105 kilo-meters from Ruhengeri to Kigali, together with thepossible installation of some hydroelectric capacity.The supply rate would average 520 liters per second(45,000 cubic meters per day), which would certainlysatisfy the current demand for water in Kigali.

Piping water this way would represent a substantialproject, with an estimated cost in the region ofUS$117 million (excluding power generation).Thus, itmight provide a good opportunity for private financ-ing and operation through a suitable BOT scheme.Theprospects for the success of such a scheme are increasedby the fact that the people of Kigali already pay a highprice for water.Thus, it is likely that the scheme can bemade financially attractive to a private operator with-out requiring an increase in the general level of tariffs.

If the Ruhengeri-to-Kigali pipeline scheme istaken forward, it should be done on an integrated basisas part of a broader basin master plan. It is in any caseunlikely that Kigali will benefit from increased suppliesfrom this source for about 8 to 10 years.There is, how-ever, an urgent need to improve supplies to the citywithin a much shorter time span. In part, the securityof supply to Kigali can be improved through a sus-tained program to reduce leakage from the distributionsystem—a key aim of the ELG management contract.Although necessarily a matter of speculation until aprogram to reduce unaccounted-for water is put inplace, it seems likely that a high proportion of Kigali’sunaccounted-for water is attributable to administrativerather than physical losses.Thus, even very substantialreductions in Kigali’s unaccounted-for water will notsignificantly increase the effective supply of water tothe city. If this proves to be the case, priority will needto be given to improving water availability in the shortto medium term.

A pilot scheme to draw water from boreholes onthe banks of the River Nyabarongo has been largelysuccessful—although there are important lessons to belearned concerning future improvements to this typeof resource development—and offers a model thatcould be replicated to provide additional water in therelatively short term.

Urban SanitationThe density of housing development in Kigali, partic-ularly in the spontaneous informal settlements that aregrowing in the city’s periurban fringe, presents an im-mense challenge to the establishment of satisfactorydomestic sanitation. The typically small plot size andthe difficult topography mean that individual on-sitesanitation is often infeasible. Facilities for the disposalof the contents of latrines and septic tanks are inade-quate, and there is little provision for the safe removalof liquid waste or overflows. Basic night-soil removalservices do not appear to be available. Similar circum-stances apply in most other major urban centers, al-though to a lesser extent than in Kigali. A significantand growing proportion of the urban population,therefore, lacks access to adequate sanitation. Reticu-lated service is limited to a few well-off areas, and theextent of wastewater treatment is effectively negligible.An unacceptably high proportion of human waste re-turns untreated directly to the environment.

Rural Water Supply and SanitationThe proportion of the rural population benefitingfrom a main household supply or yard tap is negligible,with virtually the entire rural population relying oncommunal water sources (wells, springs, and stand-pipes). In many parts of the country, water is relativelyeasily and freely available from unprotected anduntested sources, including streams, lakes, and swamps.Hence, even where safe water has been provided, thelocal population will often use these alternative sourcesto meet at least some of their water needs. Clearly, thefirst priority is that rural inhabitants be encouraged andenabled to use safe sources of water for human con-sumption and culinary purposes.The source of waterused for personal hygiene and washing clothes is of lessconcern. Nevertheless, the use of unsafe water even forthese less important purposes can carry health risks,and the time costs for women and children engaged in


50

collecting and using water from more distant sourcesshould also be considered.

A number of factors may dissuade the rural com-munity from using potable water where it has beenprovided:• The long distances that rural people have to travel

to access safe water. In many cases, the distances areso great that they amount effectively to the com-plete unavailability of potable water.

• The lack of public awareness of the benefits of cleanwater supply, combined with a perception thatwater is not an economic good whose provisionmust be paid for.These factors continue to inhibitdemand-led provision of service.

• The price that rural people have to pay. Currentlycharges at source appear to range from FRw 5 toFRw 15 for a 20-liter jerry can (of which, aboutFRw 3 will normally be retained as payment by theperson engaged to look after the water supply pointand to recover charges).There are two distinct approaches adopted to pay-

ment for rural water service provision, in order toensure that the community can continue to meet op-erating and maintenance costs. First, a flat rate chargemay be collected by the community from all thoseconsidered to benefit from the water supply. The ad-vantages of this approach—low collection costs andperceived fairness—need to be set against the disad-vantages of the ease of evading payment and the diffi-culty of preventing (or unwillingness to prevent) thosewho have not paid from taking water. Second, usercharges may be levied in accordance with the amountof water actually taken.

The highest prices are likely to arise where signifi-cant pumping is involved in the provision of waterthrough more complex (and generally superior) sys-tems. Lack of access to electricity means that ruralschemes must generally rely on diesel pumps, and therising price of fuel is a serious concern.Ability to payis a significant issue for poorer rural communities and aprice level of FRw 10 or more can be a serious disin-centive to many to use clean water to meet even theirmost basic needs.

In addition to the price at source as noted above,people who live farther away from safe water sourcesor who are unable to carry water over the distances in-volved are likely to pay a premium for safe water that isdistributed by bicycle carriers or other vendors. It may

be noted that these water vendors, together with theindividuals employed to take care of water sources andstandpipes, are evidence of the existing low-level PSPthat already exists in the rural water sector.

Experience in the Ngenda PSP scheme in the southof Rwanda has demonstrated the following:• Willingness to pay for clean water may be very low

in rural areas where (unsafe) alternatives are readilyavailable.

• Only about 30 percent of potential consumers useNgenda water (at FRw 14 per 20-liter jerry can),and average consumption is only about 4 to 5 litersper person per day—despite the fact that standpipesare located very conveniently for the vast majorityof the local population.These findings highlight two issues:

• The design standards that should be used in ruralwater schemes. Actual overall average per capitaconsumption at Ngenda is in the region of1–2 liters per day, but the plant design was based ona consumption level of 10 liters per person per day(a general standard of 20 liters per person per day isfrequently quoted).

• The importance of effective public awareness cam-paigns in order to build demand for safe water.Boththe promotion of public health and the exploitationof the potential for PSP to make a contribution torural water supply depend on building demand forsafe water.It is estimated that about 90 percent of rural resi-

dents have access to some form of toilet facility butthat only a tiny proportion, about 10 percent, have ac-cess to an adequately hygienic facility.The majority ofpublic places and buildings in rural areas, such as mar-kets, schools, and prisons, also lack adequate sanitationfacilities. The unplanned location of dwellings, oftenon the sides of steep slopes, typically makes the instal-lation of shared sanitation facilities technically and fi-nancially infeasible.


There are several opportunities for PSP in the waterand sanitation sector.

Enhancing Standpipe BusinessesThe price that the urban poor pay for water to meettheir essential needs is high, and it would clearly be


51

desirable to find means to reduce the extra costs thatthey have to bear. While there is no control on theprices that standpipe operators charge, regulationwould be difficult to implement, and it is uncertainwhether regulation would bring any significant benefitto the urban poor. Particularly in urban areas, house-holds are normally able to choose between a number ofdifferent standpipe locations (at different levels of con-venience), and competition between operators shouldensure that prices are kept in check. A more significantfactor influencing the high premium is likely to be thefact that the operators must earn enough to providethem with an adequate livelihood. Standpipe operatorsin Kigali are earning a gross margin of about FRw6 per jerry can and—in the absence of any ancillarysource of income—will need to resell on the order of400 20-liter jerry cans daily to generate a US$1 percapita per day income (for a family of five).

It follows that a more promising approach to re-ducing the extra cost of water to the poor may be totry to develop opportunities for standpipe operators toaugment their incomes by providing additional serv-ices. A number of options have been suggested in thisregard. For example, washstands, small-scale retail busi-nesses, and even Internet cafés could be incorporatedinto extended kiosk premises.Although well intended,it is not clear that this type of development would besuccessful in most locations. Standpipe operatorswould not generally have a great deal of time to devoteto additional activities, and extra security measureswould be necessary if materials left on site are to beprotected. Such measures are unlikely to result in ageneral reduction in the premium that standpipe oper-ators charge, however, since there are relatively few lo-cations where such operations are likely to be viable.Nevertheless, a pilot project would be worthwhile totest the feasibility of this approach and to refine it.

Developing a Networked SupplyA further longer-term concern is whether a path can befound by which the urban population in general can ob-tain access to piped household water supply. At present,lack of purchasing power combined with rapid urban-ization and the unplanned spontaneous development ofthe periurban fringe presents a formidable barrier tothe provision of conventional mains supplies to thesecommunities. A possible approach would be to encour-

age very small business enterprises to establish small-scale local networks, using overground installations ofplastic piping to provide low-cost and low-capacitydistribution networks to households living within theneighborhood of an existing standpipe facility or otherbulk off-take point established for this purpose. Suchcondominial systems are becoming increasingly com-mon in Latin America and have also been introducedin Southeast Asia. Condominial systems may be devel-oped by entrepreneurs acting on their own or throughpartnership initiatives involving the government, theprivate sector, and the community.

The fact that such systems would be established atstandards that fall well below those to which ELG ad-heres is not a concern in this type of approach (espe-cially in view of the fact that systems do not yet existfor the safe disposal of wastewater). Regulation of suchschemes should be minimal, limited to occasional in-spection of the installation to ensure that basic publichealth standards are being maintained. No economicregulation should be needed, because customers wouldbe protected from monopoly exploitation by theircontinuing access to standpipe facilities.

Condominial schemes would not be feasiblethroughout all urban areas in Rwanda in the shortterm and, in particular, cannot be expected to work indistricts where the poorest urban citizens are concen-trated.Nevertheless, the approach offers the prospect ofan evolutionary path toward the eventual developmentof a universal piped water supply in urban areas and, atthe same time, should contribute to improved liveli-hoods for people involved directly in service provision.

Essential to the practicability of the condominialapproach are these conditions:• Absolute clarity that such local distribution

schemes would not conflict with any exclusiverights granted to ELG. (There are doubts about thisunder the terms of the ELG management contract.)

• Absolute clarity that it is permissible under Rwan-dan law for private persons or companies to ownwater supply assets. (It is understood that this is infact the case.)

• Confidence that rights-of-way issues could bemanaged, for example, by some form of licensing orindemnification provision.

• Development of regulated bulk supply arrange-ments to apply to ELG to ensure that condominialsuppliers have access to adequate supplies.


52

• Availability of loan financing to enable condo-minial suppliers to invest in system development.

• Provision of technical assistance to assist in theplanning and development of such schemes.Given, on the one hand, the significant potential to

improve water supplies in urban areas that the approachwould appear to have in principle and, on the other, therange of issues to be addressed in its implementation,there would seem to be a very strong case for setting upa pilot project for the condominial approach.

Private Sector Participation in Rural Water Supply and SanitationThe government’s overall policy on rural water supplyand sanitation is broadly in line with expert opinionon how these areas should be addressed, in particularwith respect to its emphasis on decentralization andcommunity participation in the design, development,and sustainable operation of demand-driven schemes.The budget currently available for the implementa-tion of specific policy measures is, however, limited,and the rate of progress of the Rural Water and Sani-tation Project is extremely slow in comparison withthe scale of need. The current pilot projects beingrun by the Department of Water and Sanitation haveindicated a marked reluctance on behalf of the com-munity to contribute toward the capital costs of infra-structure provision in advance of the provision ofservices (10 percent of the initial capital cost is gener-ally requested).The sanitation component of the proj-ect is currently limited to sensitization campaigns,with no direct help being provided for infrastructuredevelopment.

Although opportunities for PSP in rural water andsanitation are likely to be very limited in scale, theremay be scope to increase the number of individualsearning a living (or, at least, supplementing their in-come) from the provision of water services or from an-cillary services, and to enhance the level of incomesthey are able to generate.

To date, a number of national nongovernmental or-ganizations (NGOs) appear to have been more success-ful than the government in promoting community-leddevelopment of rural water services. Some nationalNGOs have also been directly involved in assistinghouseholders in installing satisfactory basic latrinefacilities. The relative success of the NGOs is almost

certainly attributable to the fact that they operate closeto local communities and are much better able to sen-sitize and mobilize the community than are govern-ment authorities.Although statistics are not available, itseems likely that the NGOs are also able to obtain abetter return from scarce resources than is the govern-ment. It may be questioned whether the government’scurrent approach takes sufficient advantage of thescope to cooperate with the NGOs in the process ofcommunity-led infrastructure provision.

The current policy approach is demand led only in-sofar as the design and planning of schemes is commu-nity driven. It remains top-down and supply driven,however, insofar as the selection of districts for inclu-sion within the scope of rural water and sanitation ini-tiatives remains a matter for the government and aidagencies. A better outcome might be possible with amore highly demand-driven approach, under whichcommunity groups, in coordination with NGOs,might be empowered to take a greater role in the se-lection of projects.

A possible model for a more demand-driven ap-proach is the approach developed in Mozambique forfunding rural electrification schemes. This model hassix main features:1. Establishing a central rural water supply fund using

government and donor contributions2. Allowing bids for subsidy from the fund to be

made—on a project-by-project basis—by any or-ganization engaged in the development of ruralwater supply projects, including the private sector,NGOs, and community groups

3. Allocating available funds to water supply projectson the basis of the total capital subsidy required peradditional person supplied (that is, favoring projectsin which unit cost is low or community or privatesector capital contributions are high)

4. Mandating that all projects be supported by plansand clear community commitments that confirmtheir long-term financial and operational sustain-ability

5. Establishing preset limits on the maximum propor-tion of annual funding that can be allotted to anyindividual project

6. Developing a closely collaborative approach withlocal NGOs in order to support their core role insensitizing and mobilizing rural communities and in


53

helping communities identify their needs and de-velop and implement project plans.A key feature of this type of approach is the need to

develop a flexible stance on standards of provision.Thepoor will benefit from better water supplies only ifthe policy emphasis is on promoting their ability toachieve improvements that they can afford, rather thanon establishing rigid minimum requirements that maybe beyond their means.

Adopting a strongly demand-led approach to ruralwater supply—in close cooperation with donors andNGOs—should lead to a step change in the rate atwhich improved water supplies are made available tothe rural poor and will maximize the benefit obtainedfrom scarce government and donor resources.

Opportunities also exist to exploit the potential tosupply some rural areas from ELG’s network. In anumber of areas, ELG’s water mains run close by ruralcommunities that lie outside ELG’s defined sphere ofoperations but which it would be practicable for ELGto serve if an appropriate business, financial, and regu-latory model could be developed. In terms of a generalbusiness model, partnership arrangements with a localprivate business or community association wouldseem to represent the most appropriate approach.ELG would enter into a bulk (that is, wholesale)supply agreement with the local operator, who wouldthen be responsible for all infrastructure, services,and revenue recovery beyond the point of meteredoff-take from the ELG system.This model is, in fact,very similar to the approach discussed above in rela-tion to extending reticulated water services in urbanareas.

Issues

Several issues arise with respect to the water and sani-tation sector.

Water Resource ManagementThe focus of this report is on infrastructure issues relat-ing to the provision of water supply and sanitationservices to the Rwandan public rather than on waterresource management. Nevertheless, the way in whichRwanda manages its water resources has critical impli-cations for water supply and sanitation and vice versa.

The “National Water Resources Management Policy”of July 1998 notes the following main water resourceproblems:• Insufficient technical professional staff• Increasing industrial pollution and lack of adequate

wastewater treatment facilities• Incomplete inventory of water resources• Lack of watercourse development and develop-

ment of potentially navigable lakes• Inadequate watershed management, high levels of

soil erosion, and increasing instability of flowvolumes

• Application of agricultural technologies that areunsuited to water resources management

• Lack of framework legislation and regulations gov-erning water planning and management in the in-dustrial sector

• Pollution of watercourses and lakes, especially bywater hyacinth.These issues will be addressed under the legal and

institutional component of the newly awarded JapanPolicy and Human Resources Development Fundgrant aimed at preparing a new project related to na-tional water resource management.

Some of these issues bear particularly directly onwater supply provision.As noted in appendix A, the in-creasing turbidity of its existing water sources is caus-ing considerable operational problems, as well as addi-tional expense, for ELG. The lack of a satisfactoryinventory of water resources is, likewise, a concern inextending the availability of potable water in ruralareas. Although water resource management policiesare clear, there is currently relatively little evidence ofsuccessful implementation, which is likely to rely onsignificant external support and cannot be expected toattract private funding.

Urban Water SupplyTwo issues arise with respect to urban water supply: theELG management contract and the role of the Multi-sector Regulatory Agency (MSR).

Management Contract The success of the ELG manage-ment contract will be of central importance to thefuture development of potable water supplies inRwanda’s urban areas. The future ability of the com-pany to tap private sources of financing, whether


54

through the intended route of a long-term concessionor through an alternative means such as bond financing,hangs critically on its ability to improve its operatingand commercial performance and to generate cash sur-pluses to remunerate private investors.

It is understood that agreement has been reachedon the terms of the ELG management contract, al-though the contractor is not yet in place.This agree-ment represents a clearly positive development for theurban water sector.The particular terms of the contractremained private at the time of drafting this report, soit is not appropriate to comment on them here. In gen-eral, however, it is clear that, if the long-term objectivesof the government are to be realized, the performanceregime must provide a framework of incentives thatstrongly encourages the management contractor toimprove revenues and to minimize costs, while achiev-ing satisfactory standards of service. For these incen-tives to be effective in practice, it will also be necessaryto have a strong, equitable mechanism for monitoringand compliance in place.

Regulation The role of the MSR is critical in the latterregard. The regulator is required both to implementsector regulations and to serve as an impartial arbiterfor dispute resolution, although it is proposed that theBoard of Directors of ELG should undertake these re-sponsibilities until the MSR is fully operational. Thegeneral enabling legal framework for the MSR hasbeen established, and funding is in place.The MSR hasmade key senior appointments, including the manag-ing director and a number of departmental heads.More junior staff members remain to be appointed,and attention will need to be given to training andprovision of technical assistance before the MSR willbe capable of performing its functions. The proposedwater sector law has not yet received parliamentary ap-proval, however, and is still in the drafting process.Thenew law is necessary to define the specific duties andpowers of the MSR with respect to urban water serv-ices and is essential to complete the overall regulatoryframework within which the management contractwill be performed.

Urban SanitationAlthough the policy objectives of the government aresensible, the policy measures established to put them

into effect are extremely limited and the overall ap-proach can be regarded as, in essence, a statement of as-pirations. The proposed Sanitation Code has not yetbeen finalized. Proposed legislation on sanitation iscontained in both the draft Sanitation Act (emanatingfrom the Ministry of Health) and in the draft Environ-mental Management and Protection Act, and there is asignificant risk of inconsistency between the two laws.This issue will also be addressed under the legal andinstitutional component of the newly awarded JapanPolicy and Human Resources Development Fundgrant.

The environmental and health consequences of in-adequate urban sanitation are extremely serious and areincreasing with growth in the urban population. It isimperative that the government and the relevant cityand town authorities address urban sanitation servicesas a matter of highest urgency, establishing clear, realis-tic, and measurable policy objectives and effectivemeans for their achievement.

Sanitation provision does not typically represent asuitable target for large-scale PSP in low-income coun-tries. Even in the capital, where average incomes arehighest, it is unlikely that a significant proportion of thepopulation would be willing to pay full cost-recoverycharges for anything more than basic on-site sanitationprovision.The sort of communal sanitation schemes thatare likely to be required to provide adequate servicesto many urban communities in Rwanda are unlikelyto represent a suitable vehicle for private financing,although a contribution by the community towardcapital and operating costs may be both possible anddesirable.

The private sector is, however, already active in thecollection and disposal of sludge from septic tanks inKigali, a service that is also provided by the Kigali CityAuthority. This activity is primarily focused on wealth-ier communities, where residents are both more likelyto have septic tanks and more likely to be willing andable to pay the high prices charged for pumping outand disposing of sludge. It would be in the interests ofcustomers if competition could be promoted in thistype of activity.

In the long term, the prospects for developingproper urban sanitation services in Kigali and otherurban areas appear to hang on the possibility of estab-lishing a significant income stream from which to


55

support the development and operation of sanitationservices.There appear to be two principal options:1. Attaching an additional sanitation charge to exist-

ing water supply charges, an approach that will beboth fair and practicable only when the majority ofthe urban population has access to piped householdwater supplies

2. Introducing a hypothecated property tax, necessar-ily associated with broader changes to the basis oflocal taxation and the introduction of a residen-tial property tax system—again, an approach thatis unlikely to be feasible in the short to mediumterm.

Rural Water Supply and SanitationIssues regarding rural water supply and sanitation in-clude ELG participation in water supply and the needfor public education.

ELG Participation in Rural Water Supply ELG is not cur-rently able to participate in the provision of rural watersupply even though its mains systems sometimes tra-verse rural areas that could benefit significantly fromaccess to improved water supplies.A particular issue ishow ELG might benefit from a share in the subsidiesthat are generally required to establish rural water serv-ice. Under the funding model for rural water supplyoutlined above, it should be relatively simple to find away for ELG to benefit. As long as ELG is allowed tobid projects into the fund (separately or in associationwith private sector or community partners), it will beable to access subsidies on the same basis as any otherparty proposing to provide a new rural water supply.This approach would provide a framework that wouldpromote achievement of the best value from availablefunding, adopting ELG rural solutions where theywork best.

A clearly thought-out legal and regulatory frame-work would be required to support this approach. Is-sues to cover will include the following:• Control of bulk supply charges• The obligations and standards that ELG should ful-

fill in relation to bulk supplies or the requirementto establish a service-level agreement between ELGand local rural operators

• The right, if any, of local operators or communitiesto requisition bulk supplies from ELG and the

corresponding limitations on ELG’s obligations toprovide wholesale supplies

• The monitoring and general procedures thatshould be applied.

Importance of Continuing Public Education and Communica-tion It is widely recognized that a key requirement inrelation to rural water supply and sanitation is to buildpublic demand—and willingness to pay—for an im-proved standard of service. PSP in rural water supplyand sanitation—for example, through the type ofscheme outlined above—will be feasible only if backedup by an energetic and focused communications pro-gram aimed at sensitizing the rural population to theimportance of clean water and the risks associated withusing unsafe supplies.

Recommendations

This report recommends the following actions:• Urgent attention should be given to developing a

clear, long-term policy framework for both thewater and the sanitation sectors, setting out in eachcase (and for both urban and rural areas) not onlythe objectives that the government proposes topursue but also the specific methods that it intendsto adopt in order to achieve those objectives. It isrecommended that this policy be published in apolicy statement and be widely disseminated.

• Following completion of the policy statement, anew water law should be drafted, as soon as practi-cably possible, to ensure that a proper enablingframework is put in place for the policy methods tobe adopted. For the same reason, it is also recom-mended that other laws such as the EnvironmentalManagement and Protection Act and the SanitationCode be amended as necessary (assuming that theyhave been enacted).

• A detailed feasibility and financial study should beundertaken of the scope for developing a PSPscheme (most probably a BOT scheme) for theRuhengeri-Kigali water supply project.The termsof reference for this study should include— Completion of any further engineering studies

and outline costings necessary to complete theproject; for example, the costs of modifying


56

the existing network to cater to the new sourceof supply and the costs of disconnecting anddecommissioning other water resources thatwould be made redundant by the project

— An economic and commercial appraisal ofthe project, including risk analyses from thestandpoint of the BOT contractor, ELG, andcustomers.

An environmental impact assessment should also becommissioned. Assuming that the prospects for thescheme appear to be positive, based on a general assess-ment of these studies, a plan should be put in place tocomplete the PSP project. It should include a detailedcommercial, regulatory, and contractual design, and theproject should be bid out on an open and competitivebasis.• Following an initial assessment of the scope for pro-

viding additional effective supplies to Kigali bymeans of an effective leakage reduction program, ifappropriate, a scheme should be created and evalu-ated to develop further boreholes on the banks ofthe Nyabarongo.The scheme should be developedon a PSP basis if possible, using a BOT (build, op-erate and transfer) or BOO (build, operate, andown) approach.

• A pilot project should be implemented to test thescope for urban standpipe operators to augmenttheir incomes by providing additional services atwater supply kiosks.

• A pilot project should be implemented by the De-partment of Water and Sanitation in conjunctionwith ELG to test the practicability of the condo-minial approach.Before carrying out the pilot proj-ect, the department should commission a full strate-gic and feasibility study to establish a provisionalinstitutional, legal, technical, and commercial designfor condominial schemes in Rwanda.

• A fully demand-led approach to rural water supplyshould be designed and implemented, incorporat-

ing competitive bidding for subsidies from a ruralwater supply fund. This project should be under-taken in close cooperation with donors and withboth national and international NGOs.

• A specific component of the scheme should be theintroduction of arrangements to ensure that ELG isable to participate in the provision of rural waterservices where doing so represents an efficient andeffective solution.Although water resource management is not di-

rectly concerned with the public service infrastructurerequired to deliver water and sanitation services to thegeneral population, it is extremely important that ef-fective policy means be identified and implemented topursue better management of Rwanda’s water resourceendowment and to achieve the objectives set out in theNational Water Resources Management Policy of1998. Water resource management is not an activitythat provides extensive opportunities for PSP, and inRwanda, as elsewhere, a government departmentor agency would normally undertake this activity(although specific works may be contracted out toprivate contractors). Because river basin boundariesrarely, if ever, correspond with local government polit-ical boundaries, and because of the important interna-tional issues that arise where watercourses and lakes areshared across international boundaries, water resourcemanagement is very poorly suited to decentralizedmanagement. It is, therefore, important that the institu-tional approach adopted for water resource manage-ment in Rwanda be nationally based, although riverbasin subdivisions could be established within thisbroad framework.

Note

1. These statistical indicators, all of which are derived from datasubmitted by the utilities, should be treated with some caution.Nevertheless, the picture presented of ELG is consistently poor.

57

This short chapter considers a number of Electrogaz(ELG) operational issues that have a bearing on boththe electricity and the water parts of ELG’s business.

Management and Human Resources

Skills development and knowledge transfer constitutean important element of the ELG management con-tract, as it is recognized that both management capac-ity and technical expertise are presently in short supplyat ELG. Internal communication is also seriously defi-cient. To its credit, the senior management team atELG is enthusiastic and is keen to embrace this capac-ity development.

Commercial and Financial Management

By ELG’s own admission, accurate information con-cerning its financial state is in short supply.The latestannual report available when this Country FrameworkReport was prepared (the annual report for 1999) in-dicated that the company was not financially viable andhad accumulated substantial debts. Its precarious finan-cial state is a consequence of a number of factors:• High operational costs. Costs are especially high at

the distribution level.• Extremely poor billing and collection performance.

The billing rate stood at 66 percent and the collec-tion rate stood at 63 percent in 1998. Performancehas improved a little since then, but not much.Thispoor performance results from a lack of systems and

technical skills, a shortage of human resources, theft,inaccurate meter reading (including corruptionamong meter-reading employees), and inadequateinformation concerning customer connections.The relatively recent introduction of local customercenters around Kigali (where the majority of ELG’scustomers reside) has improved matters.

• Inadequate accounting and customer managementsystems.Such systems make it very difficult to eitherunderstand or address the problem and also preventthe introduction of any kind of performance man-agement program to help improve the situation.

• The single, flat-rate electricity tariff. A tariff ofFRw 42 per kilowatt-hour was set by the Ministryof Infrastructure in 1998 and has not been changedsince to reflect costs. A formal tariff study is neededto establish economical tariffs (including variablerates) that will promote the efficient use of electri-city by customers.

Management Contract

Without doubt, the impending management contractrepresents a positive and important step to help ELGimprove its operational and financial performance.Contract agreement is also a necessary preconditionfor the short- and long-term donor funding that isvitally needed to introduce financial and managementsystems and to rehabilitate assets. Finally, the contractrepresents an important first step toward a deeper formof private sector participation.

5 Common Issues in the Electricityand Water Sectors

Recommendations

Assuming the ELG management contract is success-fully established, the following are recommended:• In relation to the provision of further donor fund-

ing during the term of the management contract,consideration should be given to introducing a per-formance program that links the level of paymentto the contractor to its success in implementing the

investment program and in achieving the program’score objectives, in accordance with the principles ofoutput-based aid.

• Following the success of the Cashpower prepay-ment program, sufficient funding should be allo-cated to this program to ensure that ELG’s plans toextend the program through the acquisition of newmeters can be realized.


58

59

Telecommunications development plays a central rolein the government’s vision of economic developmentled by information and communication technology(ICT). Unless there is rapid and sustained expansion ofRwanda’s telephone network, which is currently oneof the least extensive in Africa, it is clear that this visionwill be impossible to realize. Privatization and liberal-ization of telecommunications have been shown innumerous countries at a variety of stages of economicdevelopment to represent the core requirement for gal-vanizing rapid increases in access to modern telephonyservices and quick improvement in the range and qual-ity of services offered to consumers.

Broadening access to telecommunications and ex-tending the range of quality of available services canbe expected to support economic development in awide variety of ways. In urban districts, high-qualitytelecommunications networks offer the scope to de-velop information-intensive economic activities. Inrural areas, better access to up-to-date market informa-tion and improved contact with suppliers and cus-tomers increase the ability of farmers to maximize thevalue of their output and avoid wastage resulting fromextended periods of storage and inefficient harvestingof crops. More generally, improved telecommunica-tions will improve the ability of the community to par-ticipate in civil society and in the political life of thecountry.


Figure 6.1 summarizes key roles within the fixed andmobile telecommunications subsectors as they are cur-rently organized and managed in Rwanda. It indicatesthe changes that will arise on completion of theplanned sale of 51 percent of Rwandatel, the state-owned fixed-line operator, to a foreign strategic in-vestor, 43 percent to local investors, and 5 percent toemployees (the government will retain a 1 percentgolden share in order to maintain the power to vetocertain decisions).1

Figure 6.1 clearly shows the transition that is beingimplemented from an old-style state-monopolytelecommunications provider to a modern ownershipand competition model.The first stage of this processwas initiated in 1998 when MTN (Mobile TelephoneNetworks) Rwandacell was permitted to establish amobile telephone business in (partial) competitionwith Rwandatel. Rwandacell’s ownership structure in-volves a 46 percent holding by Tristar, a local invest-ment vehicle; a 26 percent holding by MTN; and a28 percent holding by Rwandatel.The proposed sale ofRwandatel, the latter’s disinvestment in Rwandacell,and the opening up of the market to competitionshould complete this transitional process.

The government has set up an ICT Commission,headed by the president, and has adopted a national

6 The Telecommunications Sector

ICT policy, committing it to transforming itself froman essentially agrarian economy to a knowledge-basedsociety within 20 years. The Rwanda InformationTechnology Authority has been established to assist inthe implementation of this strategy.

There are currently five Internet service providersin Rwanda, of which three—Rwandatel, the NationalUniversity of Rwanda (NUR), and the Kigali Instituteof Science and Technology (KIST)—are well estab-lished and two (Mediapost and DIN) are recent en-trants. Both NUR and KIST obtain their internationalbandwidth through the use of very small aperture ter-minal (VSAT) dishes, installed as part of the U.S.Agency for International Development (USAID)

Leland Initiative.About 60 Internet cafés are currentlylocated throughout the country, of which most are inthe Kigali and Butare areas. In the Maraba district,how-ever, local farmers have been using the Internet to lookup coffee prices and communicate with customers,demonstrating the potential that ICT has to contributeto economic growth and development in rural areas.


Core policy objectives for the telecommunicationssector are as follows:• To achieve a reduction in the waiting lists for the

installation of fixed lines


60

Figure 6.1 Telecommunications: Roles and Responsibilities

Policymaking

Planning

Regulation

Ownership

Funding

Operation

Centralgovernment




Overall policy responsibilityrests with the Ministry ofInfrastructure, Departmentof Communications.

Economic regulation is currently the responsibilityof the government (Ministry of Infrastructure).Responsibility is to be transferred to the MultisectorRegulatory Agency once it has established sufficientcapacity to accept the role.

Funding responsibility to be100% private in the futurebut with USF support foruniversal service.

Operation to be entirely the responsibility of theprivate sector following privatization of Rwandatel.

Rwandacell to becomewholly privately owned afterRwandatel shareholding isunwound.

Rwandatel is currently responsible for fixed-lineplanning. This responsibility will pass to theprivate sector on privatization.

Majority ownership of the fixed-line system willtransfer to the private sector, although thegovernment will retain a 49% share.


The Telecommunications Sector

61

• To extend telephone service to areas of Rwandathat are currently unserved or underserved

• To ensure better quality of service (fewer faults perline, quicker repair)

• To rebalance telephone tariffs so that prices arebrought more closely into line with costs

• To provide a wider variety of services, includingmore widespread Internet access

• To ensure the availability, wherever possible, of achoice of supplier

• To ensure affordable access to telephone services forRwandans in all regions

• To provide opportunities wherever possible forRwandan entrepreneurs and investors.Liberalization of the telecommunications sector

and the privatization of Rwandatel lie at the heart ofthe government’s approach to achieving its objectives.The enabling framework for this approach has beenput in place by the Telecommunications Law (Law No.44/2001) and Law No. 39/2001, which establishes theMultisector Regulatory Agency (MSR).The principalprovisions of the Telecommunications Law include thefollowing:• Establishment of a licensing framework for

telecommunications, covering both standard andindividual licenses

• Issuance of individual licenses by a governmentminister on the advice of the regulator (applicablewhere operators are providing telecommunicationsservices to the general public, where they requirerights of access to land, where they are required toprovide certain mandatory public services,or wheredefense and security issues arise)

• Issuance of standard licenses by the MSR• Regulation of the sector by the MSR, including

— Enforcement of licenses and imposition ofpenalties (which may be substantial)

— Modification of licenses— Price and quality control— Establishment and enforcement of technical

standards• Provision for the courts to take an appellate role in

the event of a regulatory decision being subject todispute

• Provision for management of the radio spectrumand associated licensing by the MSR

• Establishment of competition in the markets forinstallation and maintenance of equipment oncustomer premises

• Provision for directory and inquiry services• Provision for the establishment of a universal serv-

ice fund to cover the additional costs to be metby certain operators with universal service respon-sibilities

• Provision for mandatory interconnection oftelecommunications operators on regulated terms.

In addition, the MSR will be responsible for• Ensuring that operators maintain a satisfactory

system for dealing with customer complaints, in-cluding a compensation mechanism if stipulatedstandards are not met

• Ensuring that operators regularly publish, on acommon basis, how well they have performedagainst stipulated standards

• Administering the national numbering plan.The government is proceeding with the sale of a

majority stake in Rwandatel,with advisers in place anddue-diligence work now largely completed. Negotia-tions are under way to unwind the existing cross-shareholding between Rwandatel and Rwandacell.

Extent of Telecommunications NetworksFigure 6.2 shows the extent and basic network archi-tecture of Rwandatel and Rwandacell.

Rwandatel Rwanda’s mountainous terrain makes it bestsuited to the construction of telecommunicationsnetworks based on radio and cellular technology. Theexisting microwave transmission network is based es-sentially on the layout that existed before the genocide.The same towers are used, but radio equipment hasbeen replaced with modern units from Lucent, Harris,and Nortel. The transmission network has a mixtureof SDH (synchronous digital hierarchy) links (at155 Mbps) and PDH (plesiochronous digital hierarchy)links (at 34 Mbps). Most high-speed multiplexed linksin Rwandatel’s network use digital microwave, becausecopper cabling and optical fiber installations require agreater number of installation personnel in a countrywhere personnel are scarce and the terrain is difficult.

At the moment, Rwandatel is equipped with twomain digital exchanges, both located in Kigali: one


62

Figure 6.2 Telecommunications Networks in Rwanda

KigaliJari

MugogoRuhengiri Byumba

Nyagatare

Rwamagana

Gisenyi

Cyangugu

Gikongoro

Tumba

KibungoKanzenze

Gitarama

KarongiKibuye

New linksOld linksTo be installed

Butare

Nyarushishi

Rub

enge

ra

A. Rwandatel network

CONGO,DEM. REP. OF

BURUNDI

TANZANIAUGANDA

Rwankuba

Bugarama

Mudaso

mwa

B. Rwandacell network

[12] Sake[18] Karibu

Rwaza [6][4] Nkamira

Link to Kashongati(Uganada)

MugogoByumba [4]

Nyagatare

Karongi Repeater

Nyamasheke [4]Gikongoro [2]

[6] Huye Butare [4]

Tumba [4]Link to Ngozi

(Burundi)

Cyangugu 2 [12]

CyanguguMururu [8][16] Camp TV

2° 30'

2° 45'28° 45' 29° 00'

2° 15'

2° 00'

1° 45'

1° 30'

Gisenyi

Ngororero

Butare

Nyabisindu[2]

[6] Ruhango

Gitarama [4]

Kigali [6] Gasogi

Mount Jari Gahini [4]

Rwankuba Repeater[6] Kayonza

Rwamagana [3]

Kibungo [2]

1° 15'

29° 15' 30° 00' 30° 15' 30° 30' 30° 45'

Kagitumba [2]

Nyarupfubire [4]

Nyamata [3]

Nyabisindu

29° 45'29° 30'

Kibuye [4]Kibuye

Gisenyi 2[11]

Umuganda[9]

[18] Gisenyi

Gitarama

Ruhengeri [4]Ruhengeri

Cyangugu 1 [12]

Gikongoro


63

Figure 6.2 (Continued)

C. Rwandacell coverage

RuhengeriRuhengeriRuhengeri

Gisenyi

Ngororero

Kibuye

Cyangugu GikongoroButare

Nyabisindu

Gitarama

Source: Rwandatel and Rwandacell.

Alcatel E10B with a capacity of 10,000 lines and oneNortel DMS100 with a capacity of 100,000 lines.Traf-fic from the regions is switched by means of remoteconcentrators, meaning that all switching is done inKigali. As and when satisfied demand builds up, therewill be a case for installing local switches in locationssuch as Butare,Gisenyi,Ruhengeri, and Gitarama.Sys-tem expansion is currently taking place using wirelesslocal loop (WLL) technology. Each WLL radio subsys-tem allows for 16 E1 trunks or 480 PCM (pulse codemodulation) lines.

Rwandacell The transmission network layout of Rwan-dacell is similar to that of Rwandatel, with mostmicrowave antenna towers being shared by the twocompanies. In Kigali, Butare, and other highly popu-lated areas, Rwandacell is building additional base sta-tions in order to meet the rapidly growing demand forcellular service.The topology of Rwandacell’s switch-ing network also follows that of Rwandatel. One mainmobile switching center, an Ericsson AXE-10, is

located in Kigali and colocated with Rwandatel’s cen-tral switching facilities.

Rwandacell is investing in network expansion andin strengthening and support systems. Its coverage ex-tends to about 15 percent of the country. By way ofcomparison, in Uganda mobile telephone networkcoverage is 15 to 20 percent.

Rural Telecommunications Rwandatel’s rural network ismainly based on Alcatel’s Rurtel equipment. An un-usually large number of repeaters is required to reachthe terminal stations, reflecting the difficulties causedby the terrain. It is expected that WLL technology willbe used increasingly in the future to replace existingrural telephone systems. As yet Rwandatel has not,however, programmed the installation of any WLL sys-tems in rural areas.

Rwandatel has entered into an agreement withArtel Communications, under which the latter is en-gaged in installing solar-powered VSAT stations at 400rural locations. Each of these units will support a

PC–local area network connection and up to six tele-phone channels, which will provide public access tobasic voice and data services using a prepaid (scratchcard) system. With the completion of this projectduring 2003, Rwanda has one of the highest densitiesof public satellite telephone lines per square kilometerin the world (one public telephone per 30 squarekilometers).

Sector Performance

Rwandatel’s performance has been poor, consideredacross a wide range of service attributes. Comparedwith other African telecommunications serviceproviders, Rwandatel’s market penetration is excep-tionally low,with only 0.27 fixed lines installed per 100people.As figure 6.3 shows, this number is among thelowest in Africa. It is also clear that mobile penetrationrelative to fixed-line penetration is among the higheston the continent.The waiting list for a fixed-line con-nection is currently the highest in Africa when meas-ured as a proportion of fixed lines installed, while as apercentage of population it is the highest in Africa withthe exception of Kenya (see figure 6.4). The overallpicture presented is one of an underperforming fixed-line operator that is failing to provide adequate accessto telecommunications services—with the high levelof unserved demand being reflected in an unusuallyhigh level of mobile penetration.

The relatively high level of mobile penetrationmeans that Rwanda’s overall combined teledensity—at1.1 connections per 100 people—compares a littlemore favorably with the experience of other Africancountries, although overall teledensity remains ex-tremely low by any measure.The increasingly impor-tant role of wireless telecommunications (both fixedand mobile) services in providing voice telephonyservices in Rwanda, as elsewhere in Africa, calls intoquestion whether it is any longer appropriate to draw adistinction between conventional fixed-network serv-ice provision and wireless service provision in framingpolicy and regulatory requirements. A more relevantdistinction than whether the network technology useswired or wireless connection is probably whether thesystem that is in place supports the provision to cus-tomers of voice services, data services, or both.

Figure 6.5 shows that Rwandatel’s tariffs are gener-ally modest in comparison with those of other Africanfixed-line operators, with both line rental and chargesper national call minute being among the lowest. Con-nection charges are, however, higher than the Africannorm.

Low disposable incomes are clearly one reason whyteledensity is relatively low in Rwanda, but they arenot enough on their own to explain Rwandatel’s poorperformance. As previously noted, waiting lists areunusually long and, as figure 6.6 shows, low gross do-mestic product (GDP) per capita does not appear to


64

Figure 6.3 National Teledensity

A. National teledensity

0

5

10

15

20

25Connections (per 100 people)

Côte

d'Ivo

ire

Ghana

Mozam

bique

Kenya

Nigeria

Rwanda

Sout

h Afri

ca

Sene

gal

Tanza

nia

Ugand

a

Zambia

Côte

d'Ivo

ire

Ghana

Mozam

bique

Kenya

Nigeria

Rwanda

Sout

h Afri

ca

Sene

gal

Tanza

nia

Ugand

a

Zambia

B. Mobile compared with fixed-line penetration

0

1.0

0.5

1.5

2.5

2.0

3.5

3.0

4.5

4.0

Mobile phones (per fixed connection)

Fixed-line teledensityMobile teledensity

Sources: BMI-Techknowledge 2002; International Telecommunications Union.


65

have held back fixed-line penetration to the same ex-tent in other African countries.

Factors Underlying Rwandatel’s Poor PerformanceA number of factors explain Rwandatel’s poorperformance:• The genocide resulted in a very serious attrition of

senior managerial and technical staff and com-pletely undermined the company’s knowledge base,including information regarding the location of itsown network.

• Of the company’s approximately 320 employees,only about 4 percent have higher-level education.The total number of employees is itself low byAfrican standards, when compared with the num-ber of lines provided. Rather than indicating a highlevel of efficiency, this low level of employmentprobably reflects Rwandatel’s inability to recruitand retain appropriately qualified personnel. As aresult, Rwandatel lacks adequate capacity to man-age the business effectively and is forced to concen-trate on “fighting fires” rather than on improving

Figure 6.4 Fixed-Line Waiting Lists

A. Fixed-line waiting list

0

10

20

30

40

50

60Waiting list (percent of fixed lines)

Côte

d'Ivo

ire

Ghana

Mozam

bique

Kenya

Rwanda

Sout

h Afri

ca

Sene

gal

Tanza

nia

Zambia

Côte

d'Ivo

ire

Ghana

Mozam

bique

Kenya

Rwanda

Sout

h Afri

ca

Sene

gal

Tanza

nia

Zambia

B. Fixed-line waiting list

0

0.04

0.02

0.06

0.10

0.08

0.14

0.12

0.18

0.16

Waiting list (percent of population)


Figure 6.5 Fixed-Line Tariffs

A. Residential monthly line rental

0

2

4

6

8

10

12US dollars

Côte

d'Ivo

ireKen

ya

Mozam

bique

Rwanda

Sout

h Afri

ca

Sene

gal

Tanza

nia

Ugand

a

Zambia

B. Charge per call minute

0

0.05

0.15

0.10

0.25

0.20

US dollars

Côte

d'Ivo

ire

Sout

h Afri

ca

Ghana

Mozam

bique

Kenya

Nigeria

Rwanda

Sene

gal

Tanza

nia

Ugand

a

Zambia


the quality of operations and management planningand reporting.

• The very high rate of unplanned housing develop-ment since 1994, combined with the lack ofknowledge about the location of network assets, hasmeant that Rwandatel is often unable to obtain ac-cess to its core assets.

• The company is failing to develop plans formedium- and long-term development and has beenoperating on a six-month to one-year planninghorizon—probably reflecting concerns that any in-vestment now may conflict with the plans of a newstrategic shareholder. Rather than take the initiativein planning, the company has been excessivelyswayed by proposals generated by potential suppliers.

Demand GrowthA forecasting exercise carried out by the InternationalTelecommunications Union in collaboration withRwandatel in 2001 predicted that if demand were to befully satisfied there would be a need to move from acombined teledensity of 0.3 lines per 100 people in2001 to 1.4 in 2006 (representing 112,000 fixed-lineand 13,000 mobile connections for a population of

8,907,000). In fact, by the end of 2001, there were al-ready some 69,000 mobile subscribers—as comparedwith about 22,000 fixed-line subscribers—representinga combined teledensity of 1.1. As previously noted,take-up of mobile connections has been unexpectedlyrapid, leaving growth in fixed-line subscriptions wellbehind.


Opportunities for private sector participation (PSP)exist in both telecommunications and ICT.

TelecommunicationsThe most significant opportunities for PSP in telecom-munications are those that emerge directly from theliberalization of the sector and the partial privatizationof Rwandatel, namely for strategic investment inRwandatel itself and for the entry of more privatelyowned operators into the telecommunications market.It can be expected that new entrants will focus on theprovision of wireless services, providing both fixed andmobile services through wireless technologies.

Information and Communication Technology There are further opportunities for PSP in ICT, partic-ularly in the context of Rwanda’s strategy of ICT-leddevelopment.Most of these opportunities are relativelysmall in scale and particularly well suited to exploita-tion by the country’s indigenous businesses. Examplesof such opportunities include the following:• The provision and operation of Internet cafés and

rural and urban telecenters (or public data process-ing centers),where the community can access com-puter facilities and the Internet

• The provision of computer and computingsupport, training, and related services to thegovernment.The extent to which these opportunities can be re-

alized will depend in part on the development of im-proved telecommunications facilities—the bandwidthavailable to Internet cafés is currently limited, forexample—and on public awareness and educationprograms that will generate demand for ICT serviceprovision.

In the case of telecenters, there are also a number ofdifferent PSP models that may be adopted, potentially


66

Figure 6.6 Teledensity and Per Capita Gross DomesticProduct

0

1.00

0.50

2.00

1.50

2.50

3.00

3.50

5.00 3500

3000

2500

2000

1500

1000

500

0

Connections (per 100 people) Per capita GDP (US$)

4.00

4.50

Côte

d'Ivo

ire

Sout

h Afri

ca

Ghana

Mozam

bique

Kenya

Nigeria

Rwanda

Sene

gal

Tanza

nia

Ugand

a

Zambia

Fixed-line teledensity Per capita GDP (US$)



67

operating alongside one another:• Phone shops, probably offering only basic tele-

phony and facsimile services and generally operatedas individually owned microenterprises or as part ofa national telecommunications provider franchise

• Basic telecenters offering telephony, fax, basic com-puting, and Internet access services and ownedor franchised by small and medium-size privateenterprises

• Multipurpose telecenters providing access tohigher-end technologies and additional trainingand specialized services.These centers will generallyrequire full-time expert staffing and PSP is likely torequire the involvement of larger and more sophis-ticated businesses with significant financial capacity.

Issues

The most important issue in relation to the telecom-munications sector is whether the government’spolicies and the effectiveness with which they areimplemented will be sufficient to stimulate the devel-opment of effective competition in the market andpromote the extension of services to a significantlygreater proportion of the community. It is generally ac-cepted, on the basis of extensive international experi-ence, that effective competition provides by far themost potent stimulus available to accelerate improve-ments in service coverage and quality of service.

Telecommunications LawA number of decrees required to make the Telecom-munications Law and the related terms of the law es-tablishing the MSR fully effective have not yet beendrafted.These decrees include the following:• A decree or decrees covering the amount and the

basis of fees to be paid by license holders in order tosupport the funding of the MSR

• An interconnection decree, setting out the termsand the pricing principles to be incorporated in theinterconnection agreement

• A licensing decree or decrees, setting out thescope of activities for which a license is required,the nature of the license to apply to different net-work and service activities, and the conditions andprinciples that will be incorporated in differentlicenses

• A decree setting out establishing a universal accessfund and determining the basis on which contribu-tions will be made to this fund.These decrees must be addressed as a matter of ur-

gency, because the regulatory regime cannot functioneffectively without them. Equally important, in theabsence of these decrees the prospects of a successfulprivatization of Rwandatel will be significantlyreduced.

InterconnectionEven if the issue of the interconnection decree is re-solved, considerable additional work is likely to be re-quired, in particular by the MSR, before it is ready todecide on the specific issues raised in any interconnec-tion dispute regarding terms and pricing.The terms onwhich mobile operators terminate calls that originateon competitors’ networks will be of considerable im-portance to how competition develops, as will be theinterconnection terms set by Rwandatel.

Establishing a framework that allows free and faircompetition in the market for international calls willbe especially important. Particular attention must begiven to• The terms on which operators terminate incoming

calls that originate on competing operators’ inter-national networks

• The provisions made to allow subscribers on partic-ular networks to select a preferred internationalcarrier to complete outgoing international calls.Setting appropriate interconnection terms and

prices is a complex task, raising difficult theoretical andpractical issues.Tackling this task will represent a majorchallenge for the MSR, at a time when it is also en-gaged in many other important tasks necessary to es-tablish the regulatory regime and has had relativelylittle opportunity to build up its experience and ex-pertise. That said, the issue of interconnection is onethat has now been addressed in many countries and the(publicly available) interconnection agreements thathave been established in other developing countriesprovide an excellent starting point for the preparationof a suitable agreement in Rwanda.

Issuance of LicensesA further important concern is the potential for ex-tended delay in the issuance of licenses to applicants

wishing to enter the market. Under the Telecommuni-cations Law it is the responsibility of the MSR to ad-vise ministers on whether the government should issueindividual licenses to telecommunications operators.At the time this report was prepared, there was a queueof three applicants for mobile licenses. Neither theMSR nor the Ministry of Infrastructure is at present ina position to issue the requested licenses.There is a se-rious risk that, by the time a new license or licenses areissued, Rwandacell’s penetration of the mobile marketwill have reached the point where it will be very diffi-cult for a new entrant to compete on fair and equalterms.

The MSR is preparing to commission a consul-tancy study concerning the number of mobile opera-tors that the Rwandan market can sustain, and WorldBank funding has been allocated for this purpose. It isimportant that this study progress as quickly as possibleand that it does not further delay the licensing of newentrants.

There is a lack of clarity about how the govern-ment and the MSR intend to proceed if the market isconsidered to be too small to accommodate all the cur-rent and potential entrants. The generally acceptedapproach is to auction the available licenses. TheTelecommunications Law makes provision for the useof tendering approaches, however, only where thegrant of an individual license involves the use of ascarce resource (for example, the frequency spectrum).It does not appear to provide for a tendering process incircumstances where the government is seeking tolimit the number of licenses to be issued for commer-cial or economic reasons.

Moreover, any such auction would deplete the re-sources of the successful bidders and would further re-strict their ability to compete on equal terms withRwandacell. It is further understood that Rwandacellitself was not required to make any payment in returnfor the 10-year mobile telephony concession that itholds, and it is expected that Rwandatel will (in linewith usual practice) be issued a mobile license to im-prove the prospects for its successful privatization. Ob-viously these two factors complicate the process of de-signing a fair, efficient, and lawful means of selectinglicensees.

The current policy is to issue separate and distinctfixed-line and mobile licenses. As already noted, tech-

nological and market developments mean that it is nowhighly questionable whether it is appropriate to main-tain this separation. A preferable approach would beto issue standard licenses that would not place anyrequirement on licensees in terms of the particulartechnology that they should use, although they mightstill impose obligations (where appropriate) on theservices to be provided and incorporate rollout targetsfor service coverage.

Independence of the Multisector Regulatory AgencyThe Telecommunications Law reserves importantpowers to the government, in particular with respect tothe issuance of individual licenses. Although the gov-ernment is required to seek guidance from the MSRon such matters, there is a clear reduction in the inde-pendence of the regulator and a risk that politicalconsiderations might interfere with the process of issu-ing licenses.Broader concerns regarding the independ-ence of the MSR are discussed in chapter 7 in relationto cross-cutting issues.

Rural TelecommunicationsAccess to telecommunications can bring substantialbenefits to the rural population. It offers farmers theopportunity to market their crops more effectively andaids participation in civil society.Where rural commu-nities fall within the scope of mobile telephone net-works, simple kiosk facilities or shared prepaid mobiletelephones can give relatively poor individuals someaccess to telecommunications.

Very low incomes and relatively low populationdensity, combined with the unavailability or unreliabil-ity of local electricity supplies and Rwanda’s moun-tainous terrain, mean that for the foreseeable future itwill remain uneconomical to install either fixed-line ormobile telecommunications infrastructure in manyrural districts. Artel is working under the terms ofRwandatel’s license and has not itself been issued alicense. Hence, it is unable to set up VSAT installationsin, for example, private business premises. Because ofthe particular nature of the market in which Artel isseeking to operate, the issuance of a license to Artel isunlikely, other things being equal, to result in a sub-stantial diminution of the market available to otheroperators.


68


69

There are, however, concerns that the close con-tractual relationship between Rwandatel and Artel mayprejudice the opportunity for free and fair competitionto develop in telecommunications, because other oper-ators will be unable to obtain access to Rwandatel’snetwork on equal terms.Although there is, in general,a strong case for issuing a standard license to Artelwithout delay, doing so should be associated with theestablishment of an interconnection regime applicableto all operators on an equal basis and correspondingadjustments to the contractual arrangements betweenthe company and Rwandatel.

Recommendations

The overall policy that is being pursued in relation tothe telecommunications sector is well founded. Themain recommendations in relation to the sector aretherefore concerned with improving the effectivenesswith which this policy can be implemented.The recommendations are as follows:• Urgent attention should be given to drafting the

important decrees that are required to put theTelecommunications Law into effect and that areessential both to the privatization of Rwandatel andto the development of the regulation and competi-tion regime.

• Technical assistance should be procured as a matterof urgency for the head of the telecommunicationsdepartment in the MSR in order to increase the

pace at which key regulatory rules and mechanismscan be developed.At the same time, such expertisewill help the department head build his knowledgeabout complex aspects of telecommunications reg-ulations, notably interconnection.

• Documents should be drafted as soon as possiblethat define the way that the government intends toexercise its functions under the Telecommunica-tions Law. The aim of the documents should beboth to define and to confine the circumstances inwhich the government might consider acting otherthan as specifically advised by the MSR.

• The study to assess the appropriate number of par-ticipants in the Rwandan telecommunications mar-ket should proceed as quickly as possible, and li-cense documents should be drafted in parallel withthis process (and in advance of the issue of the li-censing decree). It is suggested that a model basedon the issue of standard (technology-independent)licenses be adopted.

• Urgent attention should be given to the develop-ment of a fair, efficient, and legally permissiblemeans of allocating available licenses to applicants.

Note

1. Rwandatel SA was created as a separate company from Rwanda’spostal services in January 1993, with a 99 percent governmentshareholding. The remaining 1 percent is held by six otherRwandan companies and organizations.

Some of the biggest constraints in the development ofRwanda’s infrastructure—and, in particular, on the useof private sector participation (PSP) to improve infra-structure services—are not specific to the individualinfrastructure sectors but act across a range of sectors.These cross-cutting issues may also have an importanteffect on private sector development in the widereconomy. This section explores the main cross-cuttingissues identified in the course of the Country Frame-work Report (CFR) exercise and, where possible, pro-poses possible actions to remediate them.

Institutional Capacity

The institutional capacity of both government andparastatal entities is a core issue that must be addressed.

Government The ability of Rwanda’s government institutions tomeet the challenge of reforming and extending thecountry’s key infrastructure sectors and to promote therole of PSP in contributing to this process is threatenedby a number of factors impinging on the institutionalcapacity of the government. Some of these factors areamong those normally to be expected in a developingcountry where resources are scarce, government isfaced on all sides by pressing needs for action, andminimal indigenous experience exists in relationto designing and implementing reform policies forinfrastructure.

The government’s institutional capacity is alsothreatened by a number of factors that are specific toRwanda, including the legacy of the genocide, and theneed to adjust to the changes brought about by decen-tralization.

The capacity of government institutions at both thenational and local levels was seriously damaged by thegenocide.The loss of key experienced staff members, inparticular, but also the destruction of records and theinterruption of programs and projects established be-fore 1994 have diminished the ability of the govern-ment to perform as effectively as would otherwise bethe case.While there are many highly competent gov-ernment officers, there remains a lack of strength indepth. In addition, some of the senior staff members donot have a continuous track record of experience ingovernment that might have been expected were it notfor the disruption caused by the genocide. Institutionalmemories may be shorter than would normally be thecase. These direct effects of the genocide are com-pounded by indirect effects, including the need to di-vert scarce resources toward national rehabilitation,resettlement, and reconciliation efforts.

In the long term, Rwanda’s policy of decentralizingkey government activities, initiated in 2001, shouldstrengthen the government’s ability to deliver infra-structure sector reform and infrastructure developmentprojects at the local level. It should, in particular, lead toa closer focus on the least well-served groups, whichwill often constitute the poorest and most vulnerable

70

7 Cross-Cutting Issues


71

in society. In the short term, however, the transfer offunctions and resources to district levels has inevitablybeen disruptive in the following ways:• The district authorities need to build their capacity

to deliver on the government’s infrastructure poli-cies but may lack the technical skills necessary tomanage larger, more complex projects, such as thedevelopment and operation of local electricity andwater networks.

• The central government has also lost the ability tocarry out some of its responsibilities through thetransfer of budgets, staff, and equipment to the dis-tricts. Within the Ministry of Infrastructure, for ex-ample, the Department of Water and Sanitation iscurrently unable to monitor water resource avail-ability and quality.Among the more general institutional capacity is-

sues that the government faces are the need to establisha commission to monitor and combat corruption; andthe need to improve financial systems and accountabil-ity within the government. Key priorities include therationalization of the functions of the Auditor Gen-eral’s Office and the Inspectorate of Finance and Audit,the establishment of a legal framework for the NationalTender Board, and the wider dissemination of budget-ary and financial performance information.

Parastatal OrganizationsMany of the effects of the genocide on government in-stitutional capacity are mirrored in Electrogaz andRwandatel, two important utility organizations. Bothorganizations not only suffered serious damage to theirinfrastructure and records but also lost numerous keymanagers and the knowledge and skills that they pos-sessed. A high proportion of the current senior man-agement of these parastatal organizations was recruitedafter 1994 and accordingly does not have the depth ofexperience and understanding that would usually befound in long-established utilities. Although much hasbeen achieved in terms of rebuilding the capacity ofboth Rwandatel and Electrogaz since 1994, muchwork still remains. In both organizations, however,there are capable and dedicated managers who areaware of the weaknesses in their company’s perform-ance and who have worked unceasingly to offer thebest service they can to the community.

In addition to management issues, several other fac-tors have an important bearing on the capacity of theparastatals. These include an exceptionally high pro-portion of temporary staff, poor training facilities, in-adequate accounting systems, and poor commercialand customer management capabilities.

RecommendationsInstitutional capacity issues cannot be resolvedovernight, and simply importing expertise from out-side Rwanda will not on its own generate a sustainableimprovement in institutional capability. Imported ex-pertise may also result in the development of ap-proaches that are out of tune with the underlyingculture and customs of the country.The following ap-proaches are recommended for supporting a programto develop PSP in the infrastructure sectors:• Survey the resources that are available in each of the

key departmental units that have core responsibilityfor infrastructure development and operations, fo-cusing in particular on the availability of skilled andexperienced personnel in key positions.

• If the CFR recommendations are accepted in eachinfrastructure sector, have each key departmentalunit carry out a survey of additional resources, per-sonnel, and equipment needed to carry out thoserecommendations. Combine the estimates fromeach key departmental unit into an estimate for theministry or parastatal concerned.

• Develop and execute a training plan, with prioritygiven to areas of departmental operation that arecentral to the design and implementation of infra-structure development programs and PSP in infra-structure.

• Ensure that a core element of the training providedis in the form of technical assistance programs andthat such programs include clear, measurable targetsin terms of the knowledge transfer to be achieved.Alongside capacity building focused on training

and staff development, it is recommended that admin-istrative load be made a key criterion in evaluating, pri-oritizing, and developing infrastructure projects, in-cluding PSP. Lack of capacity in depth means that aheavy administrative burden falls on a narrow spectrumof staff members in both the government and theparastatals. Therefore, it is vital to avoid unnecessary


72

complexity in areas such as PSP scheme design inorder to ensure that available resources are exploitedeffectively.

Local Financial Markets

The following issues arise in relation to the capacityand operation of local financial markets:• Lack of a local, in-depth, medium- and long-term

bank debt market to finance private sector infra-structure projects.The current level of bank interestrates is far too high and needs to be reduced to aneconomic level.

• Lack of a local capital market to provide opportuni-ties for raising private equity by floating the equityof private companies investing in or operating pub-lic infrastructure services.

• Lack of sufficient local microfinance facilities to as-sist local entrepreneurs in establishing small-scaleinfrastructure facilities, such as small water treat-ment and supply facilities or small rural hydro irri-gation and power generation facilities.

• Lack of local bank expertise, knowledge, and expe-rience in project finance techniques and securitystructures based on project cash flows.The CFR recommends that the government of

Rwanda focus on the following areas to enhance localfinancial markets:• Encourage and support the local banks in providing

medium- and long-term bank debt and financingfor company start-ups. This effort will require thedevelopment of a medium- to long-term bank de-posit market, initially created in the corporate sec-tor, particularly by insurance companies, which areseeking additional outlets for investing insurancepremiums. The government should assist theprocess by establishing a national pension programwith tax incentives for pension savings and withpension savings partially invested in the medium-and long-term deposit market.

• Use technical assistance and training programs totrain bank staff in medium- and long-term lendingpractices, project finance techniques, and provisionof financing for start-up companies.

• With donor assistance, provide additional microfi-nance facilities through local banks.

• With donor assistance, consider establishing a smallinformal equity trading market, in addition to thepresent underused commercial paper market, possi-bly similar to the Alternative Investment Market(AIM) in London, using donor assistance.

Regulation

An effective and well-designed regulatory frameworkwill be critical to achieving pro-poor private sectorparticipation in infrastructure in Rwanda.The designof this regulatory framework should include the fol-lowing:• An appropriate framework of controls and incentives

that promotes quality and efficiency improvementsthat extend coverage to unserved communities(generally the poorest citizens)

• Protection for service users from exploitation bymonopoly infrastructure service providers

• Conditions in which competition among infra-structure service providers and the extension ofchoices for consumers can flourish,where both fea-sible and economically efficient

• The ability for private sector participants to earn afair return that reflects the risks they are acceptingand the quality of their performance.For such a framework to function effectively, it is

vital not only that the key rules, mechanisms, and pro-cedures associated with the regulatory framework be inplace, but also that they be put into effect in a compe-tent, clear, and fair—but firm—manner by a strongregulatory institution.

The role of the regulator is effectively to “hold thering” among government, private owners and opera-tors, service users, and the unserved community. Bal-ancing the often conflicting interests of these differentgroups in a way that not only is fair but also can clearlybe seen to be fair demands that the regulatory bodyhave the highest achievable degree of independence.At the same time, the complexity of the issues that willarise requires that the technical capacity and resourcesavailable to the regulator be sufficient to match the dif-ficult challenges that will be encountered.

Considerable attention has been devoted to gettingthe regulatory approach right in Rwanda. As a result,both a well-considered regulatory law and a properly


73

funded regulatory institution are now in place. Thereare, nevertheless, some important regulatory issues tobe addressed.

Urgency of Developing Regulatory CapacityFirst, the Multisector Regulatory Agency (MSR) hasbeen established only very recently (November 2002).As of the date of this CFR, the managing director andthree sector department heads have been appointed,together with some administrative personnel. Fundingis in place for the appointment of further technical staffmembers, and the MSR is pursuing the tasks of staffplanning and recruitment as a matter of urgency. Nomatter how speedily additional staff members are ap-pointed, however, the MSR will not be fully functionaluntil training is complete. This task will be consider-able given the lack of previous experience in Rwandain the area of independent economic regulation, and,therefore, it will not be completed quickly.

At the same time, a number of vital regulatory tasksmust be carried out very soon.These include draftingkey license documents, developing an interconnectionregime in telecommunications, and developing moni-toring systems and regulatory controls for Electrogaz(without which the proposed management contractcannot function properly). There is also an urgent needfor an enhanced energy law (or separate electricity andgas acts), as well as a transport, water, and sanitation lawif the MSR is to regulate effectively.

Indeed, unless sector-specific technical assistance—in addition to the general regulatory assistance alreadyto be provided to the managing director—is providedright away to assist and guide MSR staff membersthrough these difficult tasks, the job facing them willprove overwhelming, and serious delays and mistakesare likely to be made.

Independence of the MSRSecond, under the approach set out in the regulatorylaw, executive regulatory authority rests with the Reg-ulatory Board and not with the MSR.The MSR has anadvisory and secretariat role and is not responsible forfinal regulatory decisions. This approach is not unusualand has been adopted in many other countries, includ-ing the United Kingdom, where it is used for postalsector regulation. The seven individuals who have beenappointed to the Regulatory Board by presidential

decree are, however, all members of government min-istries or closely related bodies, such as the KigaliMayor’s Office.All have clear policy responsibility for,or a direct interest in, one or more infrastructuresectors.

The individuals who have been appointed are all ofhigh standing and are well placed to understand theregulatory issues that will arise in the infrastructuresectors and the reasons underlying the chosen ap-proach. There are, however, two serious drawbacks.The most important of these is that the approach seri-ously compromises the independence of the MSR: theRegulatory Board would appear, in effect, to be an ex-tended arm of the government. This fact will in-evitably impair the prospect for successful PSP inRwanda or, at least, add to the risks perceived by po-tential investors, consequently increasing the rate ofreturn that they will require.

Furthermore, because the board members havebeen appointed on the basis of the official positionsthey hold in government organizations with infra-structure interests, senior staff movements betweenministries are likely to force changes in the composi-tion of the board relatively frequently.The RegulatoryBoard will hence be less able to build a strong commonunderstanding of regulatory principles and issues, andit will be less effective as a consequence.

Both of the issues noted above concerning boardmembers could be avoided if membership on theboard were drawn from civil society rather than fromgovernmental bodies. Ideally, legal provisions shouldmandate such an approach. In addition, the selectionprocess for board members should, if possible, be lessinfluenced by the government. For example, at leastsome of the members might be selected through pub-lic selection committees.

These concerns regarding the lack of independenceof the regulator are increased by the particular ap-proach that has been adopted by the Telecommunica-tions Law, which is the only sector-specific lawcurrently in place.As noted in chapter 6, the Telecom-munications Law gives the key regulatory functions tothe government.The State Ministry of Telecommuni-cations under the Ministry of Infrastructure retains re-sponsibility for many of the most important elementsof regulation, including issuing individual licenses tokey operators, deciding who does and who does not


74

require a license, setting quality of service parametersand associated compensation levels, and determininginterconnection pricing principles.

Development of an Appellate BodyAn additional cross-cutting issue regarding infrastruc-ture regulation in Rwanda concerns the process forappeals against regulator decisions. The current lawmakes the courts responsible for adjudicating disputes.As yet, however, no commercial court exists inRwanda, although it is understood that there are pro-posals to establish one. Even then, however, the natureof the disputes that are most likely to arise in relationto regulatory decisions will often require considerablespecial expertise in both technical and economic mat-ters if the disputes are to be satisfactorily resolved.There is also a significant risk of substantial delay anda real danger that dominant operators will be able totake advantage of this delay. For example, by disputinginterconnection terms in the courts, a dominanttelecommunications operator could easily prevent acompetitor from entering the market. By the time thecourts have adjudicated, the opportunity for profitablyentering the market may no longer exist, and competi-tion may have been stifled right from the beginning.

RecommendationsTo address the issues discussed above, the CFR recom-mends the following actions:• Reconsider the composition of the Regulatory

Board to ensure that at least the majority of mem-bers are from civil society rather than the govern-ment and that a public selection committee orequivalent arrangement be introduced to broadenthe basis on which appointments are made. Enactlegislative amendments to make this composition aformal and permanent requirement of the regula-tory system.

• Amend the regulatory law, changing the role of theRegulatory Board to a general supervisory one,with executive authority for the majority of regula-tory decisions transferred to the MSR managingdirector.

• Provide technical assistance to each of the depart-ment heads over a one- to two-year period toenable them to deal effectively with the heavilyfront-loaded work that they face and to enable

them to build their level of knowledge and exper-tise as quickly as possible. Measurable targets forknowledge transfer should be incorporated withinthe technical assistance contracts.

• Evaluate and implement effective institutional andprocedural arrangements for appeals against regula-tory decisions to ensure that the appeals process isaccessible but that it discourages frivolous chal-lenges and delaying tactics.The question of the need for transport sector regu-

lation has also been raised; however, it is fairly evidentthat the primary area that regulation must address isthat of compliance with technical specifications andsafety standards.The need for economic regulation ofthe transport sector should, thus, be of secondary con-cern to the need for technical regulation. The onlypossible exception to this rule remains with the needfor enhanced economic regulatory control in the air-ports sector.

Accounting System

Rwanda’s accounting system lacks a national account-ing standards system. In addition, the lending institu-tions lack confidence in the corporate accounting in-formation provided to them to support loan requests.The Revenue Authority also encounters problemswith tax payments, often disputing corporate tax andprofit calculations.

The CFR recommends the following actions to al-leviate accounting system issues:• Require that all corporate bodies over a certain size

prepare their accounts to minimum internationalaccounting standards.

• Ensure that all local practicing accountants haveminimum accounting qualifications.

• Set up a national accounting standards board orcommittee to regulate national accounting standards.

Taxation System

The following issues arise in relation to Rwanda’s tax-ation system:• No tax holidays exist to attract private sector in-

vestors.• The carryforward period for tax losses is only three

years.


75

• Private sector investors consider Rwanda’s tax col-lection procedures to be harsh and arbitrary; hence,they are a constraint on increasing private sectorinvestment.

• The private sector considers importation proce-dures to be cumbersome and slow, thus impedingprivate sector investment.The CFR recommends the following changes to

the current taxation system:• The government should consider a tax holiday for

investors of five years from the start of operations.• The carryforward period for tax losses should be

increased to five years.• The tax collection procedures should be revised to

make them more client friendly and less confronta-tional.

• The importation procedures should be revised toreduce delays.

Legal System

Both general issues and sectoral issues arise with re-spect to the legal system.

General IssuesRwanda is a civil law country, and as such its entirelegal structure may appear unfamiliar to those who areused to an Anglo-Saxon system of common law.Thisfact does not mean that the necessary structures andlaws are not there. It simply means that the structuresand laws may take a different form from those in coun-tries with common law systems. Whether this civil lawstructure is helpful or acceptable in terms of encourag-ing outside investment is a matter for debate. The pri-vate sector within Rwanda can be assumed to be fa-miliar with the system, as would be private investorsfrom other civil law countries. Furthermore, some as-pects of the Rwandan administrative structure mayprovide considerable support to the small-scale waterand energy projects discussed elsewhere in this report.

Rwanda’s civil laws are based on a Civil Code to-gether with local customary laws. In particular,Rwanda has inherited the French system of public ad-ministrative law, in which there is a division of theSupreme Court that is devoted to constitutional andadministrative matters. The existence of a system ofpublic law means that intérêt publique has legal status

and has a dramatic impact, for example, on the treat-ment of land ownership.

Legislation is made in the form of laws passed bythe National Assembly, by presidential decree, and byministerial decree. Presidential and ministerial decreestake the place of the secondary legislation known tocommon law systems.

A few recommendations should help foreign in-vestors in their dealings with the legal system:• One of the principal difficulties is that much of

Rwandan law is written in Kinyarwandan andFrench. Under the terms of the Arusha Accords,English is now one of the official languages of thecountry. The more recent laws are published inEnglish. It is recommended that, to assist access tooutside investors, all laws relevant to the privatiza-tion process be made available in official translationsinto English.

• Another difficulty is that the applicable laws are noteasily ascertainable to outsiders or even, in somecases, easily obtainable. This difficulty undoubtedlyarises in part because of the problems that Rwandahas experienced in the recent past. To make it eas-ier for the private sector to access applicable laws, acentral library with good cataloguing facilities, or aone-stop government publication office that canprovide information about the existing laws forparticular areas would be invaluable and would as-sist outside investors in carrying out due-diligenceexercises.

Sectoral IssuesAs far as the sectors under consideration in this reportare concerned, the principal legal issues appear asfollows:• Roads. Ownership of and liability for maintaining

roads differ according to whether they are in thepublic or private domain. This basic civil law con-cept informs all of Rwandan land laws.

• Water and sanitation. Rights to run pipes over theland of others, rights relating to sources of water atthe district level, and the legal status of local au-thorities are all important issues.

• Energy and telecommunications. In legal terms, energyand telecommunications are both affected by rightsin contracts, the legal status of parastatals, and regu-latory structures.


76

More generally, any outside private sector operatorwill want to see a transparent disputes resolution sys-tem in place, preferably both domestic and foreign; afavorable tax regime both for operation in Rwandaand for repatriation of profits; and a company lawsystem that enables private sector operators to controlany jointly formed companies in which they may beinvesting.

Land Law Certainty of land ownership will be essentialfor any large-scale private participation in infrastruc-ture.The basic premise governing Rwandan land law isthat all land belongs to the state and can be transferredonly under certain conditions. A distinction is madebetween domaine publique and domaine privé in landtenure, as in every other facet of civil law society. Landin the domaine publique can only be sold or transferredto a private individual or company if it is first trans-ferred into the domaine privé. Such transfers must bedone by legislation.

A new initiative is under way to deal with registra-tion of title to land as part of a reform of the land lawin general and to address the problem of customaryrights to the use of land, which are widely found inRwanda (see appendix B for details).

The main problem with Rwanda’s land law con-cerns the public-private ownership rules,which appearto restrict alienation of any kind—including conces-sion contracts—to land held in the private domain,even by the state. It has already been necessary in oneparticular private sector project to legislate for thetransfer of land held by a parastatal to the private sec-tor. It seems likely that this need will arise in otherprojects. Whether it is efficient in terms of legislativeand government time to legislate such matters piece-meal is questionable, to say nothing of the delay thatthe process causes.

A further problem is that Rwanda’s land law pro-vides for droit de reprendre—the right of the state to lit-erally take back land that is not being properly ex-ploited. In addition, problems with encroachmentappear to affect road building and maintenance. Theseproblems may be a result of customary law and may bedealt with by the proposed land law reforms.

Given the complexities of Rwanda’s land law andthe fact that reform is currently under consideration, itis recommended that a further study be made with aview to including in the proposed new law provisions

that resolve these problems so that land used in PSPschemes is not subjected to what are—to an investorfrom a non-civil law country—quite serious disadvan-tages and constraints.

Water Rights Water rights are significant in two ways:• The water supply responsibilities of Electrogaz and

its proposed transfer to the private sector • The rights of small private suppliers to access and

pipe water over the land of others.The water supply responsibilities are affected by the

provisions of administrative law and intérêt publique; therights to access and pipe water are affected by the spe-cific legislation on water rights and the provisions ofRwanda’s land law.

The Civil Code deals with water rights in the sec-tion on propriété, or property and ownership.The waterin streams and lakes and underground watercourses canbe owned by no one. Subject to the laws or regulationsthat govern the use of such water and concessions thatmay be awarded by the public authorities, the right touse such water is common to all. If the land in questionis in the domaine privé, the right to pipe water over theland of another is dealt with by means of a droit deservitude—a type of way leave—and appears not topresent a problem, except that a fee might be payable.

There is, however, a law (of July 23, 1979) provid-ing for expropriation of property for public use thatmight be useful in this context. The law details a pro-cedure for confiscation of private property by decree(either presidential or ministerial, depending onwhether the project is local or regional) and for thepayment of compensation.

There may be a need for minor amendments to thelaws relating to water rights, or possibly a licensingscheme, to overcome delays in making available rightsof way for private sector pipelines.

Administrative Structure An area of civil law practice thatmay prove extremely valuable in light of recommenda-tions in chapter 4 relating to small-scale water projectsis the legal and administrative rights and responsibilitiesgiven to the districts. These form the administrativebackbone of the state; they own land—in both publicand private domains—and generate and manage theirown local projects for roads, water, health, education,and so forth.The new law on local authorities, passedin 2001 following the postwar decentralization process,


77

gives districts the power specifically to work with gov-ernment and private entrepreneurs in matters con-cerning the welfare of the population, with particularreference to policies that help protect and improve theliving conditions of the poor.The district can also ac-quire shares in associations or nongovernmental organ-izations (NGOs) where it has interests and can appointits own representatives in such organizations.

The central government authority is represented bythe préfet, or governor, of the province in which therelevant district is situated.The governor is assisted by aregional committee, which has statutory responsibilityfor, among other things, coordinating projects beingcarried out in partnership with the private sector.There is, therefore, in Rwanda the legal and adminis-trative foundation for local-level projects involvingpublic and private participation and for a monitoringrole for central government, which will need to en-force common standards.

Contract Law The law governing contracts generallydoes not differ greatly from that found in common lawsystems. Some technical rules governing formation dif-fer, but these rules are unlikely to significantly affect ei-ther an outside investor of the kind contemplated hereor the small-scale nonprofit schemes recommended inthe water and energy sectors.

The Civil Code precisely details the law relating tothe different forms of contracts that govern normalcommercial life. It reflects civil law systems in defininga contract, the terms used in the contract, and partici-pants in the contract. In addition, the Civil Code dealsspecifically with several matters of significance to thisCFR, including the status of foreigners. Foreignershave the same rights under Rwandan law as aRwandan national. Contracts entered into by foreign-ers are subject to specified conflict-of-laws rules,whichstate that the status and capacity of the foreigner are tobe assessed according to the laws of his or her owncountry, unless the rules of that country are unknown.

The concept of the concession contract was devel-oped under French law, apparently as a result of the re-strictions relating to alienation of public property (dis-cussed above).The Civil Code of Rwanda contains achapter setting out the law relating to concessions, orbail emphytéotique.

Apart from the basic provisions of the Civil Code,no laws deal specifically with concession contracts;

build, operate, and transfer projects; and so forth.Thereis provision for laws (or decrees) to be passed forspecific projects, but because of the need fortransparency—and on the basis of discussion in thisreport about the road system and the need for mainte-nance and improvements by the private sector—consideration also should be given to upgrading exist-ing laws on concession contracts. If changes were madein contract laws, then the problem noted in the sectionon company law below regarding the liabilities of amanager might be able to be addressed at the sametime.

Legal Status of Parastatals Public enterprises in Rwandaare governed by the Organic Law of November 7,1975 (as amended).The law provides that a public en-terprise (établissement publique) is a public service cre-ated by law and endowed with personnalité civile and ad-ministrative and financial autonomy. The enterprisesgoverned by the Organic Law include Electrogaz, aswell as OCIR-Café and OCIR-Thé—the parastatalsset up to run the coffee and tea sectors.

Privatization of public enterprises is governed bythe Privatization Law of March 11, 1996. A presiden-tial decree on May 3, 1996, set up a Privatization Com-mission, which is charged with directing the work ofprivatization. The Privatization Law provides widepowers for the method of transfer—partial or total liq-uidation, leasing out, or restructuring—but the meansadopted are to be the means used to set up the enter-prise in the first place (usually legislation). In addition,the process is restricted to situations in which (a) themanagement of the entity is considered unprofitable,(b) the state wishes to withdraw from a commercial orindustrial concern, or (c) the purpose for which theentity was created has been attained. There is also anoverriding provision that the state retain a shareholdingwhen national sovereignty or security is concerned.These provisions obviously affect the proposed estab-lishment of a management contract for Electrogaz, de-scribed in detail in previous chapters. They will alsohave to be taken into account in privatizing thetelecommunications sector.

The restrictive nature of the Privatization Law maymake it difficult to achieve the private participationneeded in national parastatals. It is recommended thatthe law be amended to provide that, when privateparticipation is in question, rather than a wholesale


78

disposal to the private sector, a more liberal regime beadopted, particularly with regard to the parastatals thatcan be selected for privatization.

Dispute Resolution The court system in Rwanda is basedon the French model. It is headed by a Supreme Court,which has five sections: the Department of Courts andTribunals, the Court of Appeals, the ConstitutionalCourt, the Council of State, and the Revenue Court.The Supreme Court ensures the constitutionality oflaws and statutory orders before promulgation into lawand makes decisions on appeals from decisions of lowercourts and administrative authorities.While the exist-ing methods of dispute resolution may be adequate toregulate local commercial activity, even the new Arbi-tration Act—which contains provisions for public an-nouncement of decisions and allows appeals to thecourt in certain cases (including possibly intérêtpublic)—will probably not prove an acceptable tool foroutside investors.

The law applied by the courts in civil and commer-cial matters is based partly on those sections of theCivil Code dealing with commercial matters (and theso-called Commercial Code, which is composed of anumber of different statutes) and partly on local cus-tom, as required by article 98 of the Rwandan consti-tution. Otherwise, the courts apply the general princi-ples of law and equity. A system of criminal courts alsoexists, which was renovated and strengthened after thegenocide.

Civil law countries generally rely heavily on the set-tlement of disputes, particularly commercial disputes,amicably or by negotiation.This reliance is evidencedby the 1998 law establishing the Rwanda InvestmentPromotion Agency (RIPA).That law provides for theamicable, if possible, settlement of disputes between aforeign investor (as defined in the law) and RIPA or thegovernment. If an amicable settlement is not possible,there can be reference by agreement to any interna-tional body to which both the country of the investorand government of Rwanda subscribe. Alternatively,the March 1965 International Convention on Settle-ment of Disputes between States and Nationals of otherStates of March 1965 can be referenced.

Rwanda has a comparatively new arbitration law,which is largely intended to resolve local disputes.Thecountry has established an arbitration center under the

terms of a World Bank project designed to assist thedevelopment of the private sector.

However, the system of dispute resolution is some-what lacking in transparency for an outside investor.The statutory disputes resolution methods laid out inthe 1998 law establishing RIPA need amending to ac-cord with international practice. In addition, the newlocal authority laws contain provisions for dispute res-olution when districts are involved in a dispute thatinvolves reference to a minister and, therefore, to theRwandan courts. These laws may also need amend-ment if the local authority structures are to be used forlarge-scale PSP projects.

Company Law There is no Commercial Code in termsof a single piece of legislation.The law relating to com-panies, however, is set out in the Companies Act of1988.This act provides for the types of commercial or-ganizations that may be established in Rwanda.The actis based on the French model and provides for theseforms of business organization: the public company (so-ciété anonyme), the private company (société à responsabil-ité limité), and two main forms of partnerships.There isalso provision for temporary associations for specificprojects and profit-sharing associations, but these enti-ties do not have legal personality.The main models arealso found in Anglo-Saxon systems and do not inthemselves present a particular problem to an outsideinvestor.There are the usual requirements for local reg-istration, representation, and so on. Detailed provisionsof the Companies Act, however, have been found toconflict with requirements for the privatizationprocess, as discussed later in this chapter.

In addition to the Companies Act, separate legisla-tion exists for cooperatives, which have a considerablerole in Rwanda, particularly in the agricultural sector.

Two problem areas arise concerning company law.First, one way for a non-Rwandan private investor toparticipate in privatization is to form a joint companywith one or more local commercial organizations.Usually the investors would form a private company,which would effectively be a partnership between theparties.The outside investor would retain managementand legal control of its investment by keeping a major-ity of the shareholding and, either by means of a man-agement agreement or by the appointment of a man-ager, would also retain decisionmaking control.


79

A private company is usually chosen because that formof business organization affords the parties more confi-dentiality and restricts the transfer of shares to thirdparties.The Privatization Act, as reflected in the man-ual issued by the Privatization Agency (set up underthe Privatization Commission), however, requiresshares in state-owned bodies being privatized to besold by tender. This requirement appears to conflictwith the rules relating to private companies and thetransfer of their shares.

Second, the Companies Act provides for the man-ager of a private company to have statutory liability forhis actions, both to the shareholders and to the credi-tors. As mentioned above, this requirement would ap-pear to place at risk the management arrangementsadopted by many outside investors, particularly small-to medium-size investors.

There must be a structure by which an investor canprotect its need to manage its investment. It is in theinterest of the government to create a legal climate thatencourages investment.The alternatives are either to letthe investor fend for itself and choose corporate struc-tures other than the private company, since other formsof business organizations do not have the problemshighlighted above, or to include in legislation somesort of provision whereby a private company set up toparticipate in a PSP project is not bound by these pro-visions of the Companies Act. Neither alternative issatisfactory, and further examination of this problem isneeded.

Bid Evaluation Process

One issue arises in relation to the bid evaluationprocess. Although there has been some improvementlately, the current system through the National TenderBoard is still considered by the private sector to becumbersome and cause delay, which unnecessarily in-creases both bidding costs and the cost of the particularproject. It is therefore recommended that the currentsystem be reviewed and procedures be developed to re-move any unnecessary delays and reduce project costs.

Project Selection, Prioritization, and Planning

A basic procedure is in place for all public sector infra-structure projects: the Public Investment Program

(PIP). Normally, all public sector projects would needto be accepted by the ministry governing the particu-lar sector for inclusion in the PIP. These ministriesmeet annually with the Ministry of Finance and Eco-nomic Planning to agree on the PIP for the year ahead.A National Investment Strategy prioritizes sectors byvarious criteria, including, for example, what the ex-pected social impact would be, what budget resourcesare available, whether the project will be financed by agrant or loan, and whether government counterpartfunds are required. Projects, especially large infrastruc-ture projects, would also normally require a feasibilitystudy, in many cases carried out by an independentconsultant selected in competition through theNational Tender Board. The project then goes to theInvestment Council. If approved and if funds are avail-able, the project then goes to implementation by theline ministry.The Central Public Investments and Ex-ternal Finance Bureau (CEPEX) is responsible formonitoring and coordinating the full project cycle,including monitoring the identification of projects,coordination, and resource mobilization; monitor-ing the bid evaluation; and monitoring the projectimplementation.

The procedure is, however, flexible. If the projectsare emergencies or self-evident—for example, an ur-gent road rehabilitation—they are derogated from thenormal process and go straight to the InvestmentCouncil.The approval by the council is a formality. If adonor offers funds for a certain project, it will be in-cluded in the PIP and be referred to the InvestmentCouncil for approval, which also will be a formality.

The Ministry of Infrastructure is responsible for thepriority of infrastructure projects. At the end of theday, however, the Ministry of Finance and EconomicPlanning controls the available financial resources andthe coordination of the planning process.

An issue identified in the course of this CFR exer-cise is whether the flexibility in the procedure needstightening, with additional checks and balances to re-duce the possibility for misuse through, for example,undue ministerial influence, lack of donor and projectcoordination, or overriding national investment prior-ities.The government should therefore review the cur-rent procedure to ascertain whether any additionalchecks are necessary to ensure that the procedure isnot misused in any way and that the right balance is


80

maintained between needs and resources. The objec-tive is to ensure that all projects that go forward toimplementation meet the criteria of the NationalInvestment Strategy through the PIP.

Government Policy Communicationwith the General Public

A need exists for an improved presentation of govern-ment policies to the general public and greater publicsensitization to the reasons for—and benefits of—PSPin infrastructure. It is recommended that the govern-ment consider making the Ministry of Infrastructureresponsible for publicizing government policy on PSPin public infrastructure service provision.To do so, theMinistry of Infrastructure will need in-house publicrelations expertise. Alternatively, the PrivatizationAgency could be made responsible, or a new govern-ment public relations department could be establishedthat would be responsible for publicizing all govern-ment policies.

Environmental Issues

Environmental concerns act as constraints to the devel-opment of certain forms of infrastructure, and theCFR takes this into account. A more detailed analysisof the environmental situation in Rwanda and the im-portance of environmental issues in the developmentof infrastructure nationally is provided in appendix Ato this report.This section provides only a brief reviewof the key environmental issues.

Environmental Policy and LegislationRwanda’s environmental policy and legislation are cur-rently in a limited state of development. The keyframework legislation on the environment, the Envi-ronmental Management and Protection Act (EMPA), isstill in draft form.The EMPA requires further attentionto avoid duplication with other draft legislation (par-ticularly the draft Sanitation Act, which addresses bothwastewater and solid waste). Of importance, the draftEMPA does not match international requirements forthe completion of environmental impact assessments(EIAs).This is particularly relevant to the future devel-opment of infrastructure in Rwanda, because EIAs willbe a fundamental component of the process to be

followed and will ensure that environmental protectionconcerns are addressed in all major projects. Rwandaneeds guidelines for the EIA process that reflect inter-national best practice, as espoused, for example, byEuropean Commission Directive 97/11/EC. In addi-tion, the EMPA will require detailed regulations for itsimplementation, and priority must be given to theproduction and finalization of these regulations.

Strategic environmental assessments (SEAs) willalso be of use in Rwanda, especially to clarify the envi-ronmental effects of major government policies andprograms. SEAs are best formulated using EuropeanCommission Directive 2001/42/EC. The EMPA willrequire amendments to include SEAs within its scope.

Sectoral Environmental ConcernsEach of the sectors analyzed in this report has accom-panying environmental concerns.

Transport A range of environmental concerns should betaken into account in the development of the transportinfrastructure in Rwanda. The development of roads,airports, and railways all carry significant potential forgenerating adverse environmental impacts (for exam-ple, dust emissions, noise, impacts on land use, and im-pacts on biodiversity).The EIA process should be usedto minimize such adverse effects. European Commis-sion Directive 97/11/EC (and similar internationallegislation from elsewhere) can help identify projectsthat should be subjected to the EIA procedure.

Energy—Power Generation and Distribution The fact thatRwanda relies mainly on hydroelectric power impliesthat the environmental impacts from power generationare limited once the schemes have been constructed.The construction of new hydroelectric generatingfacilities, however, will have significant environmentalimpacts, and the proposed plan at Nyabarongo (in-volving a 41-meter high dam, a reservoir of 320hectares, and the displacement of some 2,150 inhabi-tants) will need to be prefaced by a full-scale EIA.Micro-hydro plans are likely to be of some relevanceto rural populations, although solar power and biogaswill also be important in some circumstances.The en-vironmental impacts associated with the distributionof electricity (including visual intrusion, exposure toelectromagnetic fields, and polychlorinated biphenyls


81

in transformers and capacitors) should be identifiedand addressed.

Energy—Methane Gas The Lake Kivu methane resourceis of great importance to the energy sector in Rwanda,and its development is a high priority.The relationshipbetween this resource and hydroelectric power gener-ation (as well as the importation of power) should beclarified through the completion of an SEA.The SEAwould also serve to consolidate the government’s pol-icy on energy generation as a whole. A full-scale EIAwill be needed on any plan that includes the exploita-tion of the Lake Kivu methane resource; in addition,the existing pilot plant will need to be replaced. AnEIA also should extend to the uses of the methanedownstream, which will involve significant environ-mental benefits (because of the replacement of fossilfuel and wood fuel use and the reduction in green-house gas emissions).

Water and Sanitation—Water Supply The current cost forpotable water in Rwanda is very high, reflecting thedifficulties in abstracting and treating (mainly) surfacewaters. The per capita water consumption in bothurban and rural areas is low, but demand will certainlyincrease as the economy develops and the populationexpands.The potable water supply in Kigali is inade-quate in relation to both quantity and quality; there-fore, the realization of the proposed Ruhengeri-Kigaliwater supply plan should be given high priority.Smaller water supply plans in the rural areas should becapable of attracting private investment, although thegovernment should maintain control over the standardsachieved in such projects.

Water and Sanitation—Sanitation Currently, there is al-most no treatment of wastewaters in Rwanda in eitherurban or rural areas.Without a doubt, this lack of watertreatment is a major driver of human health problems,particularly because the topography of the country im-plies that watercourses will be subject to multiple usesin their various locations.The introduction of twinnedcharging for water supply and sanitation appears tohave some long-term potential, but it will only be pos-sible when individual household connections, ratherthan shared supplies, predominate. In rural areas, theuse of low-technology sanitation plans is preferable.

The disposal of various types of sludge should receiveparticular attention, because it can create significantadverse environmental impacts.

Telecommunications Only two minor environmentalconcerns relate to the telecommunications sector: thevisual intrusion of overhead fixed lines (especially inareas of elevated protection status) and the possible ef-fects of electromagnetic fields on human and animalhealth. Neither of these concerns is considered partic-ularly important to the development of the telecom-munications infrastructure in Rwanda.

RecommendationsA number if recommendations can be made with re-spect to environmental issues:• Policy development.The government should expend

additional effort to clarify its environmental poli-cies, including updating the National Environmen-tal Action Plan.

• Development of legislation and guidelines. The EMPAshould be updated and improved, and duplicationwith other draft laws (for example, the SanitationAct) should be eliminated.The procedures relatingto EIAs in the EMPA should be amended to reflectinternational best practice, as exemplified by Euro-pean Commission Directive 97/11/EC. In addi-tion, SEA procedures should be cited in the EMPA.Guidelines for both EIAs and SEAs should be pro-duced. Detailed regulations under the EMPAshould be developed as a matter of priority.

• Use of EIA procedures.EIAs should be used as the keytool in identifying and mitigating any environmen-tal impacts arising from specific forms of infrastruc-ture development.The completion of EIAs will beimportant for all major transport-related infrastruc-ture development (roads, railways, and airports).European Commission Directive 97/11/EC pro-vides guidance that should be reflected by theRwandan procedures.

• Energy sector projects.The current government policyin the energy sector is unclear, and the completionof an SEA on energy generation in Rwanda is rec-ommended.Any new hydroelectric projects of sig-nificant scale should be prefaced by a full EIA. Cer-tainly an EIA is needed for the proposed project atNyabarongo (with the EIA updating and extending


82

the previous work by Sogreah).The development ofthe methane gas resource at Lake Kivu should be ahigh priority. All environmental constraints, how-ever, should be clarified through the completion ofa full EIA, and any mitigating measures should bescrupulously followed.

• Water and sanitation projects—water supply.The watersupply sector offers considerable opportunities forPSP. In urban areas, the proposed Ruhengeri-Kigaliwater supply project should be pursued with vigor.Smaller water supply projects involving PSP shouldalso be possible in the rural areas.The governmentshould specify a range of basic requirements forsuch projects, especially in relation to the requiredquality of potable waters.

• Water and sanitation projects—sanitation.The currenttreatment of wastewaters is inadequate and shouldbe improved countrywide.The population’s abilityto pay for upgraded wastewater treatment is ques-tionable, especially in rural communities. Twinnedcharging for water supply and sanitation should beconsidered, at least in urban areas.Using the Nyanzasite for the disposal of sewage sludge should be dis-continued, and an alternative disposal methodshould be identified. In rural communities, the in-troduction of improved forms of sanitation basedon low technologies (ventilation-improved latrinesand so forth) should be the primary objective. Inaddition, education on the importance of adequatesanitation should be improved.

Other Cross-Cutting Issues

Some other cross-cutting issues bear comment.

Communication between Ministries and ParastatalsA general lack of communication between ministriesand parastatals is evident.

It is recommended that communication betweenministries and parastatals be considerably improved. Ifthis recommendation is to be successfully imple-mented, it will probably be necessary for a senior man-ager in each ministry and parastatal to be responsiblefor communications between his or her ministry orparastatal and other ministries and parastatals.

CorruptionCorruption in government or within elements of theprivate sector is a major constraint on the participation

of the legitimate private sector in providing publicinfrastructure services and in investing in infrastructureassets. Many constraints exist in this area. Unfair andnontransparent bids, unfair and nontransparent evalua-tion and award of infrastructure project bids, unfair tax-ation, unfair and unreasonable application of tax laws,difficulty in obtaining any necessary approvals to pro-ceed with project implementation, difficulties in ob-taining land titles or the use of land under lease arrange-ments, difficulty in obtaining import or export licenses,and difficulty in obtaining foreign exchange and remit-ting lawfully earned dividends on equity investmentsare just some of the problems that can arise as a directresult of corruption. If corruption occurs or is even ru-mored to occur, then the legitimate private sector, bothlocal and foreign, will be reluctant or will refuse to pro-vide services or invest in Rwanda.The foreign invest-ment market is a worldwide market. Foreign private in-vestors, in particular, will go where these problems donot occur or occur to a much lesser degree.

Rwanda does not appear in the Corruption Percep-tions Index 2002 (Transparency International 2002), soit is not possible to compare the country with the lev-els of perceived corruption in other regional countries.

The local private sector business community be-lieves that corruption in Rwanda is minimal comparedwith other African countries. At most, it affects 15 per-cent of the business transacted in a given period oftime.Where it does occur, it is against the overall pol-icy of the government.The local private sector businesscommunity also believes that the government is slowlyand surely fighting corruption and that, in time, cor-ruption will be eliminated. Government actions takeninclude setting up the National Tender Board to han-dle all public procurement bids on a transparent basisand establishing the Office of the Auditor General,specifically to fight corruption. The performance ofthese institutions is good, although slow, but it willimprove once their capacities have improved.

Proposed Implementation Schedulefor the Recommendations

The government of Rwanda, in association with theWorld Bank, conducted a number of workshops thatwere funded by the Public-Private Infrastructure Advi-sory Facility and included key stakeholders, such ascivil society organizations,NGOs,donor organizations,

Cross-C

utting Issues

83

Action/project Responsibility 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012

Sector-specific action/project

Increase in all road maintenance fund charges/penalties National Roads Fund

Strictly enforce axle-load restrictions and penalize overloadedvehicles

Directorate of Roads

Prioritize roads in order of their importance to the economy Directorate of Roads

Give urgent attention to drafting decrees required to put theTelecommunications Law into effect

Department of Telecommunications

Expedite study of appropriate number of participants in the Rwandantelecommunications market; draft provisional license documents inparallel

Multisector RegulatoryAgency

Make collecting agencies responsible for passing revenue to theNational Roads Fund as quickly as possible

Ministry of Finance

Allocate funding for extension of "Cashpower" plan Ministry of Finance

Multisector RegulatoryAgency

Develop long-term policy for transport

Return control of all road maintenance to Directorate of Roads

Ministry of Infrastructure

Prepare fully costed comprehensive 10-year road rehabilitation andmaintenance plan

Ministry of LocalGovernment,Ministry of Infrastructure

Agree on financing plan for 10-year road rehabilitation plan withdonors


Ministry of Finance,Ministry of Infrastructure,donors

Develop new sector law for energy

Department of Energy

Develop long-term policy for water supply


Develop long-term policy for energy

Department of Waterand Sanitation

Develop long-term policy for sanitation Department of Waterand Sanitation

Develop new sector law for water supply Department of Waterand Sanitation

Develop new sector law for sanitation Department of Waterand Sanitation

Issue provisional licenses to Artel without delay to enable it to installVSAT equipment at private premises

Table 7.1 Implementation Schedule for Action Plan

(continued on next page)

Private Solutions for Infrastructure in Rw

anda

84

Table 7.1 (Continued)


Sector-specific action/project

Conduct feasibility studies for Ruhengiri-Kigali water transmissionproject

Department of Waterand Sanitation, Electrogaz

Develop further boreholes on the banks of the River Nyabarongo,if possible, as build, operate, and transfer (BOT) projects



Conduct feasibility study to develop plan for demand-led rural waterservice provision

Introduce arrangements to ensure Electrogaz is able to participate inprovision of rural water services


Introduce output-based aid incentives in connection with funding ofElectrogaz investment plan

Ministry of Infrastructure,donors

Procure technical assistance for the head of the TelecommunicationsDepartment in the Multisector Regulatory Agency

Multisector RegulatoryAgency, donors

Department of Telecom-munications, MultisectorRegulatory Agency

Establish 10-year concession basis for all road rehabilitation andmaintenance contracts

Establish Directorate of Roads


Consider developing a project for new taxiway and increase inrefrigeration and general storage facilities at Kanombe Airport


Rehabilitate runway and apron at Kamembe Airport

Establish Airports Authority as an independent Civil AviationAuthority



Prepare a rural energy plan, setting out a strategy for implementingrural electrification

Document the way that the government intends to exercise itsfunctions under the Telecommunications Law


Conduct feasibility study to develop scheme for demand-led ruralelectricity supply


Develop information, communications, and technology systems,resources, and capability with maximum possible private sectorinvolvement

Department ofTelecommunications

Conduct a pilot project to test the scope for urban standpipeoperators to augment their incomes by providing additionalservices at water supply kiosks


Contract Lake Kivu methane gas extraction and related powergeneration as BOT


Develop a fair, efficient, and legal means of allocating availabletelecommunications licenses to applicants

Cross-C

utting Issues

85


Update and extend the Kagera Basin Railway Study with associatedenvironmental impact assessments (EIAs)

Ministry of Infrastructure,donors

Approach the governments of the Dem. Rep. of Congo and Burundiregarding support for and interest in the Kagera Basin RailwayStudy

Pilot low-cost condominial water distribution scheme Department of Waterand Sanitation, Electrogaz

Cross-cutting action/project

Make administrative load a key criterion in the evaluation,prioritization, and development of infrastructure projects

Ministry of Finance

Make available all laws relevant to the privatization process in officialtranslations into English

Ministry of Justice

Establish primary version of all laws published in more than onelanguage

Ministry of Justice

Ministry of Finance,Ministry of Infrastructure

Develop policy to encourage banks to set up a medium- tolong-term debt market

Provide technical assistance and training programs for bank staff inmedium- and long-term lending, project finance techniques, andprovision of financing

Ministry of Finance,National Bank of Rwanda

Develop policy to encourage banks to set up a medium- tolong-term deposit market

Ministry of Finance,donors

Reconsider composition of the Regulatory Board and amend theRegulatory Law accordingly

Ministry of Finance,National Bank of Rwanda


Provide technical assistance to each of the Multisector RegulatoryAgency department heads over a period of one to two years

Have all corporate bodies over a certain size prepare their accountsto minimum international accounting standards

Ministry of InfrastructureAmend the Regulatory Law to give executive authority to theMultisector Regulatory Agency managing director

Multisectory RegulatoryAgency, donors

Set up a national accounting standards board or committee toregulate national accounting standards

Ministry of Finance

Ministry of Finance

Require all local practicing accountants to have minimumaccounting qualifications

Ministry of Finance

Consider a tax holiday for investors of five years from the start ofoperation

Ministry of Finance,Revenue Authority

Conduct survey of availability of skilled and experienced staff in keypositions in each of the key departmental units

(continued on next page)

Private Solutions for Infrastructure in Rw

anda

86

Table 7.1 (Continued)


Increase the carryforward period for tax losses to five years Ministry of Finance,Revenue Authority

Consider how investors in private sector participation (PSP) schemescan be given greater security of tenure over land

Amend laws relating to rights of way, or implement a licensingscheme, to overcome delays in providing rights of way for theprivate sector

Ministry of Justice

Ministry of Justice

Provide a way to ensure that investors are not exposed to statutoryliability for seeking to safeguard its powers of management

Ministry of Justice

Consider upgrading the existing laws on concession contracts Ministry of Justice

Consider ways to make dispute resolution more transparent;amend statutory dispute resolution methods to accord withinternational standards

Ministry of Justice,Rwanda InvestmentPromotion Agency

Address provisions for dispute resolution in local authority laws,if local authority structures are to be used for large-scalePSP projects

Ministry of LocalGovernment, Ministryof Justice

Ministry of Agriculture,Ministry of Environment

Update and improve the framework of environmental legislation andkey guidelines, and eliminate duplication with other draft laws

Use EIAs to identify and mitigate environmental impacts arising frominfrastructure development

Ministry of Agriculture,Ministry of Environment

Develop and execute training plan in areas that are central to thedesign and implementation of infrastructure development programs


Set policy for national pension plan with tax incentives for savers

Ministry of Finance,Ministry of Infrastructure,donorsMinistry of Finance,National Bank of Rwanda

Revise tax collection procedures to make them more client friendlyand less confrontational

Revise importation procedures to reduce delays


Consider the establishment of a small informal equity trading market


Review the current bid evaluation system and accelerate proceduresto remove unnecessary delays and reduce project costs


Ministry of Finance,National Tender Board

Consider making a single agency or department responsible forpublicizing government policy on PSP in infrastructure services

Ministry of Finance

Support the provision of additional microfinance facilities throughthe local banks


Clarify the government’s environmental policies and update theNational Environmental Action Plan

Cross-C

utting Issues

87


Ensure that public and private sector corruption are kept undercontrol

Establish a central facility to provide information about the law

Amend Privatization Law to provide a more liberal regime,particularly with regard to the parastatals that can be selected forPSP (as opposed to sale)

Ministry of Justice

Ministry of Justice

Consider making a senior manager in each ministry be responsiblefor communication between ministries and parastatals

Review the current project selection, prioritization, and planning procedure


Note: BOT � build, operate, and transfer; EIAs � environmental impact assessments; PSP � private sector participation; VSAT � very small aperture terminal.


88

parastatals, and core government institutions. As part ofthe review process, stakeholders worked to distill con-sensual action plans for the key recommendations aris-ing from this CFR. The resultant implementationschedule is provided in table 7.1.

No order of importance is implied in the way therecommended actions are presented; instead the Ganttchart methodology is used to draw attention to the ur-gency with which recommendations must be imple-mented if Rwanda is to continue to grow and developas an attractive locale for domestic and internationalinvestment.

The action planning charts list sector-specific ac-tion and projects, noting the allocation of responsibil-ity within the government, thus presenting clear im-plementation priorities for the government and fordonor organizations.

The effective implementation of the key recom-mendations arising from the CFR exercise will requireclear prioritization by both the government and the

donor community to ensure that the issues identifiedare addressed in a timely and effective fashion. Un-doubtedly, the allocation and acceptance of responsi-bility for the implementation of recommendations willrequire a great deal of coordination between the gov-ernment and the donor community.The relationshipsand momentum built in the course of this study mustbe harnessed if substantive change is to occur. Indeed,such coordination is likely to prove vital if pressing is-sues, such as weaknesses in the institutional capacity ofcore government institutions, are to be tackled.

The action plan in table 7.1 provides clear orienta-tion as to where resources may most profitably be de-ployed to create an effective enabling environment forprivate sector participation in Rwanda. Follow-up andreview of progress made by the government and thedonor community in the coming months and years isvital if political will is to remain firmly focused on thenecessary steps to be taken to realize the Vision 2020goals.

89

This appendix to the Country Framework Report(CFR) discusses cross-cutting environmental issues andprovides additional information on environmentalmatters of importance in Rwanda.The existing envi-ronmental regulations, policies, and legislation inRwanda are discussed initially to place the commentson sectoral issues in context.Thereafter, environmentalissues relating to the specific infrastructure sectors cov-ered by the current project are discussed.

The generic importance of environmental issuesdiffers among the various infrastructure sectors ad-dressed by this project in the following ways:• Few environmental elements of significance exist

for the telecommunications sector.• Environmental aspects pertaining to the transport

sector are of some importance, but they are onlycovered generally here because of the paucity ofspecific information on the precise infrastructuredevelopment likely to occur in this sector inRwanda.

• Significant environmental elements of relevanceexist for both the energy and the water and sanita-tion sector.Private sector participation in the development of

infrastructure is the key focus of the CFR. However,infrastructure development can occur in a rational andcontrolled fashion (with or without such participation)only if key environmental impacts are addressed di-rectly before the construction of infrastructure projectsbegin. It will be critical for Rwanda to address envi-ronmental issues specifically and in detail before and

during any attempts to upgrade the quality of the in-frastructure in the country.

Existing Environmental Institutions, Policies,and Legislation

Institutional IssuesThe institutional structure in Rwanda has not beenspecifically reviewed in this report, but comments areprovided here that are relevant to the development ofcoherent environmental controls in Rwanda.

It is important to note that the development of anyinstitutional capability addressing environmental issuesin Rwanda is relatively recent.The Department of En-vironmental Protection within the Ministry of Lands,Human Resettlement, and Environmental Protectionis staffed mainly by personnel who are relatively inex-perienced in drafting legislation and establishing co-herent systems of permitting and monitoring. Capacitybuilding will therefore be required in the future.This isan important issue for the international financial insti-tutions (IFIs) and the donor community to consider,because their financial assistance in developing infra-structure (or completing many other projects) inRwanda will be at risk if the environmental aspects arenot adequately addressed.

It is also notable that environmental concerns havenot yet been integrated into the government ofRwanda as a whole. It is not sufficient for governmentsto simply delegate environmental protection to a sin-gle ministry. Rather, high-quality environmental

Appendix AThe Environment

protection relies on the assumption of responsibilitiesof various entities across governments as a whole.Theinterface at the policy-strategy level for environmentalissues and other elements of the governmental machin-ery in Rwanda is inadequately defined to date, withopacity in the precise division of responsibilities.Thislack of clarity partly reflects the duplication of provi-sions in draft legislation referred to in the section onenvironmental legislation that follows, but it also pointsto the need for improved interministerial dialogue andplanning.

In addition, while formation of an EnvironmentalProtectionAgency has been discussed—such an agencywould be needed to implement any new frameworklegislation—there has been little substantive progress inthis regard.This lack of progress is a concern becausethe establishment of an executive body charged withimplementing environmental legislative requirementsis an urgent requirement in Rwanda.

The countrywide capability in the area of environ-mental protection is also relatively sparse at present.Thus, external assistance will be needed during theinitial period when environmental controls are estab-lished. However, it is very important that the in-country capability on environmental protection issuesbe developed as rapidly as possible. Work in this areahas implications for the need for training and otherforms of capacity building. Experience elsewhere hasshown that capability in the consulting area and amongthe main governmental bodies can be built within fiveyears or fewer, if sufficient funding is available. Rwandawill have to go through this process if environmentalcontrols of the required depth and quality are to be es-tablished in the near future.

International AgreementsThus far, the development of environmental policy andlegislation in Rwanda is minimal. Until recently, theenvironmental authorities have tended to rely heavilyon international agreements to provide a basic plat-form for developing the national approach to environ-mental controls. Rwanda has acceded to or ratified anumber of international agreements, including thefollowing:• The Vienna Convention of 1988 and the Montreal

Protocol, relating to the protection of the ozonelayer

• The United Nations Convention to Combat De-sertification of 1991

• The 1992 Convention on Biological Diversity• The United Nations Framework Convention on

Climate Change of 1992• The Stockholm Convention on Persistent Organic

Pollutants of 2002.While this action is to some extent laudable (many

of these international agreements essentially reflect bestinternational practice, at least to some degree), it alsointroduces certain problems. First, Rwanda has beenselective in acceding to these international agreements,creating a lopsided approach to environmental protec-tion. For example, as can be seen from the list of agree-ments above, Rwanda has concentrated mainly onthose that address air pollution issues and biological di-versity. In contrast, the United Nations Convention of1991 that relates to desertification is of particular im-portance in the Rwandan context, because there isgreat concern nationally over the impacts of deforesta-tion. In Rwanda, deforestation has been caused mainlyby agricultural development (often on significantslopes,with insufficient terracing), and high rates of soilerosion have resulted. As noted in a later section, soilerosion is relevant to at least two of the infrastructuresectors covered by the CFR.

International agreements almost always providemerely a basic framework of controls, with insufficientdetail. It is therefore not sufficient for Rwanda to relyon international agreements as a basis for its approachto environmental protection. Robust national legisla-tion on all environmental issues of consequence mustbe developed, enacted, and enforced.The next sectiondiscusses the current situation in this respect.

Policies and Legislation on the Environment

Environmental Policies Specific environmental policieshave not yet been formulated at the national level.Draft policies currently exist, but they remain underreview by the Department of Environmental Protec-tion and are likely to be changed in the future.The au-thorities have stated that environmental policies cannotbe promulgated until the key framework legislation ispublished. This viewpoint is flawed: the legislationshould be based specifically on the policies, rather thanthe reverse.


90

Appendix A: The Environment

91

A National Environmental Action Plan was pro-duced before 1994, but the Department of Environ-mental Protection considers it to be outdated and inneed of revision.There is currently no agreed timetablefor a revision to be completed. Hence, there is cur-rently a policy void on environmental issues inRwanda.This lack of policy is of great concern if in-frastructure development is to proceed (or indeed, ifthe environment is to be adequately protected in anyfashion).The absence of coherent and detailed policiesin most areas is a key feature of the government thatmust be addressed if significant economic developmentis to occur in the future, especially if external invest-ment is to be a key feature of such development.Thissituation has a number of important implications forthe IFIs and the donor community.

Environmental Legislation While a few preexisting lawsand decrees relating to environmental matters exist inRwanda, the key framework legislation, the Environ-ment Management and Protection Act (EMPA), re-mains in draft.The authorities estimate that it may bepassed in Parliament during 2003.

The draft version of the EMPA contains a numberof significant provisions including the following:• The right to a healthy environment is provided for.• The inclusion of the precautionary principle and

the polluter pays principle is particularly notable,because they lay a basic (and robust) platform forthe general approach to environmental protectionnationally.

• The need for sustainable development is included,which is important in providing a general frame-work for future environmental controls.

• Provisions are included on the quality of soil, water,and air; noise impacts; and waste management(although these provisions are all only of a genericnature).

• The need to maintain (and, it is hoped, increase)biodiversity is covered.

• Provisions are included that attempt to decentralizethe executive responsibility for environmentalissues.

• The link is made between land-use planning andenvironmental protection.

• A section is provided on environmental impact as-sessments (EIAs),which are of particular importance

to the development of infrastructure in Rwanda(with or without private sector participation).The EMPA constitutes framework legislation only.

Even when enacted, this framework law requires de-tailed regulations of various types for its full imple-mentation. Such regulations will take considerabletime to develop, and there has been little progress inthis regard. Hence, many of the real environmentalprotection controls are unlikely to be fully imple-mented for a considerable time.

Certain other draft legislation overlaps with theEMPA, and this issue has not yet been resolved. Thebest example of such overlap is the inclusion of provi-sions relating to solid waste in both the EMPA and thedraft Sanitation Act. A comprehensive review is neededof all proposed legislation in the environmental arenato eliminate inconsistencies and duplication and toensure that all required aspects of a coherent environ-mental protection regime have been included. Thisreview should certainly be completed before the en-actment of any of the current draft laws pertaining tothe environment.

Land-Use Planning, Environmental Impact Assessment, andStrategic Environmental Assessment Although the EMPAattempts to create a link between land-use planningand environmental protection, the current interfacebetween environmental controls and planning orbuilding controls in Rwanda is opaque.The planningand building controls are inadequate as a whole, withwide-scale illegal construction activity (especially inand around the urban centers). As a result, present gov-ernment practices continue to create future environ-mental problems because of the inappropriate juxtapo-sition of distinct forms of land use. It will be veryimportant for Rwanda to introduce strong controls onland use; this effort should be given a high priority.

With respect to the use of EIAs in controlling de-velopment, Rwanda has relied to date on statementsfrom the Rio and the Johannesburg world summits,which propose that EIAs should be completed for alldevelopment that may have a significant environmen-tal impact. Rwanda has not acceded to any interna-tional agreements regarding EIAs, despite the existenceof several such agreements (the Espoo Convention of1992, in particular, addresses transboundary issues).TheEMPA currently contains somewhat general provisions

regarding the need for EIAs, and there is no list of proj-ects or activities that would be subject to EIAs.

The Department of Environmental Protection hasstated that EIAs are currently being completed on al-most all projects.This practice, however, is not an effi-cient use of sparse resources. Rather, Rwanda shouldextend the legislation on EIA requirements to reflectbest international practice. Thus, lists of specific proj-ects requiring EIAs should be added to the EMPA be-fore its passage through Parliament. It is recommendedthat the European Commission Directive 97/11/ECbe used as the basis for best international practice andthat the Rwandan legislation reflect a similar approach.In that case, three annexes should be added to theEMPA to cover the following areas:• Projects for which an EIA is mandatory• Projects for which an EIA may be required• Relevant aspects for determining whether an EIA is

needed.Rwanda currently has no national requirement

for the completion of strategic environmental assess-ments (SEAs). SEAs are covered by recent EuropeanUnion legislation (European Commission Directive2001/42/EC) and are of particular use in determiningthe environmental impacts of government plans andprograms. SEAs are quite relevant for Rwanda, giventhe current immaturity of policy development. In atleast two sectors (see sector descriptions that follow),SEAs could be very useful in driving the developmentof robust and coherent government policies inRwanda.

Environmental Issues in the Transport Sector

Specific proposals with respect to infrastructure devel-opment in the transport sector in Rwanda are not yetavailable. As noted in box A.1, European Commis-sion Directive 97/11/EC cites several categories oftransport-related infrastructure for which the produc-tion of an EIA is mandatory before developmentbegins.The same directive includes additional types oftransport-related projects for which EIAs may berequired, depending on a number of circumstancesdiscussed in Annex III of the European CommissionDirective.

The European Union legislation is widely consid-ered to approximate best international practice.

Rwanda should therefore adopt the requirementsdescribed in box A.1. In any event, international finan-cial institutions frequently require the completion ofEIAs before they will provide grants or loans for infra-structure development projects.

The topography of Rwanda implies that certainenvironmental impacts arising from transport-relatedinfrastructure projects may be unusually severe. For ex-ample, the construction of roads and railways of signif-icant length would generate large volumes of materialfrom the cut faces of hillsides. In such scenarios, the at-tainment of a zero net balance of cut and fill materialsis advisable but could be difficult in certain projects inRwanda.

Environmental Issues in the Energy Sector

Electrical PowerBox A.2 shows the requirements for EIAs under Euro-pean Commission Directive 97/11/EC for the energysector. No large thermal power stations exist in


92

Box A.1 Requirements under European CommissionDirective 97/11/EC for the Completion ofEnvironmental Impact Assessments forTransport-Related Infrastructure Projects

Annex I, clause 7, of European Commission Directive 97/11/ECstates that environmental impact assessments (EIAs) are manda-tory in the following situations:• Construction of lines for long-distance railway traffic and of

airports with a basic runway length of 2,100 meters or more• Construction of motorways and express roads• Construction of a new road of four or more lanes—or re-

alignment or widening of an existing road of two lanes orfewer so as to provide four or more lanes—if the newroad—or the realigned or widened section of road—wouldbe 10 kilometers or more in a continuous length

In addition, according to annex II, EIAs may be required in thefollowing uses:• Construction of railways, intermodal transhipment facilities,

and intermodal terminals• Construction of airfields • Construction of roads, harbors, and port installations, in-

cluding fishing harbors• Construction of tramways, elevated and underground rail-

ways, suspended lines or similar lines of a particular type,used exclusively or mainly for passenger transport.

Source: European Commission Directive 97/11/EC.


93

Box A.2 Requirements under European CommissionDirective 97/11/EC for the Completion ofEnvironmental Impact Assessments withRespect to Infrastructure Projects in theEnergy Sector

Annex I, clause 20, of European Commission Directive 97/11/ECstates that environmental impact assessments (EIAs) are manda-tory for the following projects:• Thermal power stations and other combustion installations

with a heat output of 300 megawatts or more• Extraction of petroleum and natural gas for commercial

purposes where the amount extracted exceeds 500 metrictons per day in the case of petroleum and 500,000 cubic me-ters per day in the case of gas

• Pipelines for the transport of gas, oil, or chemicals with a di-ameter of more than 800 millimeters and a length of morethan 40 kilometers

• Construction of overhead electrical power lines with a volt-age of 220 kilovolts or more and a length of more than15 kilometers

Under annex II, EIAs may also be required for the following:• Industrial installations for the production of electricity,

steam, and hot water • Industrial installations for carrying gas, steam, and hot water

and transmission of electrical energy by overhead cables • Surface storage of natural gas• Underground storage of combustible gases• Surface storage of fossil fuels• Installations for hydroelectric energy production• Installations for the harnessing of wind power for energy

production (wind farms).Source: European Commission Directive 97/11/EC.

Rwanda, where hydroelectric power generation isthe main source of electricity. The country has only28.6 megawatts of installed capacity (as compared withan estimated demand of 40 megawatts nationally).Some electrical power, however, is imported, and plansexist to develop the extensive natural gas resource inLake Kivu (see the next section).

The proposed hydroelectric facility at Nyabarongowould undoubtedly require the completion of a fullEIA before construction begins. This project wouldprovide a further 28 megawatts of generation capacitybut would involve a dam 41 meters in height and areservoir some 320 hectares in area.The associated dis-placement of more than 2,100 inhabitants and the lossof significant agricultural land (about 80 hectares) arealso notable. A preliminary environmental report for

this was prepared by Sogreah in 1998, but this reportwould need to be updated and extended to meet a fullEIA requirement.

In the rural areas, the paucity of electrical supplies isa key driver of deforestation in Rwanda. Deforestationhas a number of consequences, and figure A.1 shows alink between deforestation, water resources, and powergeneration.The extremely high rate of deforestation inRwanda in the past (mainly to generate agriculturalland) has led to elevated rates of soil erosion, causingsignificant turbidity in local rivers and streams, even inlow-flow periods.This situation not only gives rise topoor-quality surface waters—many of which are usedas a potable resource (see the following section onwater)—but also threatens the turbines of hydroelec-tric power facilities such as the Gisengi and Gihera fa-cilities. In the absence of sufficient generating capacity,deforestation would in turn increase.What is requiredis an effort to break the vicious cycle shown in fig-ure A.1 and to establish the virtuous circle. If this effortcan be achieved, net benefits will be seen in all parts ofthe country in several sectors.

It is clear that rural electrification projects inRwanda will largely involve power sources such assolar energy, micro-hydro power, and possibly biogas.These sources all possess net environmental benefits, atleast if their introduction and maintenance is ade-quately managed. The recent introduction of solarpower projects in rural areas in many parts of Africa isparticularly notable and reflects the advances made inthe past decade in solar technology.Rwanda could cer-tainly benefit from such projects.

The Lake Kivu Methane ResourceThe Lake Kivu methane resource is considered a criti-cal element of the future development of Rwanda.Current estimates of the volume of methane availablesuggest that at least 55 billion cubic meters is present.A bilateral agreement is already in place betweenRwanda and the Democratic Republic of Congo withrespect to the future development of Lake Kivu.

As demonstrated by the incidents in Lakes Monounand Nios in Cameroon in the mid-1980s, significantrisks are associated with the exploitation of methaneresources in lake environments.Technical studies com-pleted in late 2000 on the Lake Kivu resource,however,have suggested that the risks associated with exploitation

are much lower at Lake Kivu than at the sites inCameroon, because of a more stable stratification.

Currently, two major proposals are competing forthe exploitation of the methane at Lake Kivu, and nodecision has been made on the preferred developmentoption.The scale of the existing resource,however, cer-tainly indicates the need for very careful planning inits development. Any exploitation of the Lake Kivumethane should undoubtedly be prefaced by the com-pletion of a full-scale EIA that should address not onlythe specific technology and plan to be used, but alsothe replacement of the existing pilot plant and thedownstream uses of the methane.

The Lake Kivu methane resource is particularly at-tractive from a global environmental viewpoint, be-cause the replacement of fossil fuels by methane wouldlead to reductions in greenhouse gas emissions.Methane could possibly be used in the future as a fuelsource in various industries (for example, the tea estatesand possibly the cement industry) and also as an auto-motive fuel. Such uses would have a positive environ-mental effect. Certain of the downstream uses ofmethane from Lake Kivu would reduce pressure on theforestry resource, which is of particular importance toRwanda.

There could well be scope for completion of anSEA on the entire energy sector in Rwanda: the

exploitation of the Lake Kivu resource and the furtherdevelopment of hydroelectric power appear to a largedegree to be competing possibilities, and each has dis-tinct environmental consequences. As noted in chap-ter 3, it is also vital that the government produce a co-herent national energy policy for the future. Thecompletion of an SEA would assist in the formulationof such a policy.

Polychlorinated BiphenylsPCBs are industrial chemicals used largely in the elec-trical industry (mainly as dielectric fluids in transform-ers and capacitors), and they are of very considerableenvironmental importance. International legislation onPCBs has strengthened in the past decade, because ofincreasing concerns worldwide over the extreme toxi-city and persistence of these chemicals.

There is no evidence currently available concern-ing the abundance of PCBs in electrical equipment inRwanda, but because of the geographic and temporalorigins of much of the equipment, PCBs are likely tobe present in at least some transformers and capacitors.Hence, there is a need for particularly careful attentionto the disposal of such equipment at the end of itsuseful life. There are no acceptable disposal facilitiesfor such equipment in Rwanda (or, indeed, in most ofAfrica) and the costs for appropriate disposal will be


94

A. The vicious cycle

Highrates of

deforestation

Low poweravailability

Turbidwater quality

B. The virtuous cycle

Minimalrates of

deforestation

High poweravailability

Improvedwater quality

Figure A.1 The Vicious and Virtuous Cycles Relating Deforestation to the Quality of Water



95

high.The completion of an investigation on the abun-dance of PCBs in electrical and other equipment inRwanda is recommended.

Environmental Issues in the Water Supplyand Sanitation Sector

Water SupplyIt is well documented that inadequate water quality andsanitation are responsible for widespread human healthproblems in Africa and elsewhere in the world.Rwandais certainly no exception to this general pattern. Inter-national data sources suggest that Rwanda faces partic-ular problems with respect to water supply, as evidencedby the difficulties the country has experienced in pro-viding water of acceptable quantity and quality not onlyto rural populations, but also to the main urban centers.The heavy reliance on surface waters for providingpotable supplies (in Kigali and in much of the rest of thecountry) exacerbates the difficulties, particularly be-cause the levels of water treatment are often inadequate.The monitoring of potable water supplies also does notmeet international requirements, with far too few mi-crobiological tests being completed.

The proposal to supply potable water to Kigalifrom the areas of Rubindi,Mutobo, and Mpenge in theVirunga region is of particular note.The project wouldinvolve a pipeline at least 800 millimeters in diameter,running some 105 kilometers from Ruhengeri toKigali. The vertical gradient is from 2,300 meters atRuhengeri to 1,850 meters at Mount Kigali, and nopumping would be required.The supply rate would av-erage 520 liters per second (45,000 cubic meters perday), which would certainly satisfy the current demandfor water in Kigali. A hydroelectric element to theproject is also envisaged, yielding additional minoramounts of power.

Given the poor quality of the surface water cur-rently available as a potable resource for Kigali, thisproject is of considerable interest. If managed correctly,it would essentially solve Kigali’s present water supplyproblems at a stroke and would significantly improvethe quantity and quality of water available in thecapital. However, such a project would certainly needto be prefaced by a full-scale EIA. (See box A.3 in rela-tion to the specifications of European Commission Di-rective 97/11/EC regarding water supply and sanita-tion projects.)

Box A.3 Requirements under European CommissionDirective 97/11/EC for the Completion ofEnvironmental Impact Assessments withRespect to Infrastructure Projects in theWater and Sanitation Sectors

According to annex I, clauses 11–13, and 15, of the EuropeanCommission Directive 97/11/EC, the following situations man-date an environmental impact assessment (EIA):• Groundwater abstraction or artificial groundwater recharge

schemes where the annual volume of water abstracted orrecharged is equivalent to or exceeds 10 million cubic meters

• Works for the transfer of water resources between riverbasins where this transfer aims at preventing possible short-ages of water and where the amount of water transferredexceeds 100 million cubic meters per year (transfers ofpiped drinking water are excluded)

• In all other cases, works for the transfer of water resourcesbetween river basins where the multiannual average flow ofthe basin of abstraction exceeds 2 billion cubic meters peryear and where the amount of water transferred exceeds5 percent of this flow (transfers of piped drinking water areexcluded)

• Wastewater treatment plants with a capacity exceeding150,000 population equivalent as specifically defined

• Dams and other installations designed for the holding backor permanent storage of water, where a new or additionalamount of water held back or stored exceeds 10 millioncubic meters

Annex II states that an EIA may be required for these projects:• Water management projects for agriculture, including irriga-

tion and land drainage projects• Deep drillings—in particularly drilling for water supplies• Inland-waterway construction, canalization, and flood relief

works• Dams and other installations designed to hold water or

store it on a long-term basis • Installations of long-distance aqueducts• Groundwater abstraction and artificial groundwater re-

charge schemes • Works for the transfer of water resources between river

basins • Installations for the disposal of waste • Wastewater treatment plants • Sludge-deposition sites.Source: European Commission Directive 97/11/EC.

The availability of potable water of adequate qual-ity is a problem in almost all rural areas in Rwanda.Groundwater systems are not commonly used, and re-liance on surface water implies significant threats fromupstream users. The ability of urban communities to


96

pay for higher-quality water is questionable, and at leastone pilot project (in Njenda) has encountered difficul-ties in persuading rural communities to pay for high-quality potable supplies rather than using water oflower quality collected directly from nearby sources. Itis clear that education about public health issues is asignificant requirement in such scenarios.

SanitationSanitation is perhaps the least developed of all the in-frastructure elements addressed by the current project.Even the major urban areas of Kigali and Butare haveeffectively no wastewater treatment infrastructure, as isalso the case in all rural areas.The widespread disposalof untreated wastewater to surface waters and to sub-soils gives rise to serious effects on the subsoil recipi-ents, and the health of downstream users of the surfacewaters is threatened. The collection of sludge fromseptic tanks in the capital city (which in some cases in-volves private contractors) is intermittent, and the dis-posal of the sludge occurs at the Nyanza waste disposalsite on the southern outskirts of Kigali.The disposal lo-cation is altogether improperly designed and managed,amounting to an open dumping site with no adequatecontrols on environmental impacts. There are signifi-cant problems with respect to introducing cost recov-ery in the waste disposal infrastructure in Rwanda, butenvironmental protection in this area is especiallycritical.

As previously discussed regarding environmentalpolicies, none of these areas of concern is addressed ef-fectively by the existing national environmental poli-cies or by the current legislation. No standards ofrelevance to environmental protection have been de-veloped for Rwanda (for example, there are no re-quirements for the quality of potable waters, surfacewaters, base river flows, or wastewater effluents). How-ever, the charges for water supply are currently high inRwanda compared with other developing nations.Hence, there is scope for PSP, at least in the water sup-ply sector.

Sanitation in rural areas is generally inadequate,countrywide. Efforts to upgrade latrines to so-calledventilation-improved latrines have not always met withsuccess. Inadequate sanitation, however, is a major fac-tor driving human health problems in rural areas. Solu-tions must involve low-cost technologies, because the

ability to pay for adequate sanitation is deeply suspectin rural communities.

Some of the historical projects as well as some ofthe currently proposed projects in the water and sani-tation sector in Rwanda have separated water supplyfrom sanitation. This division is contrary to the princi-ples of many of the IFIs, which generally demand thatthese elements be linked within infrastructure devel-opment projects.There is a particularly good rationalefor this requirement in Rwanda, because many of thewatercourses are subjected to multiple uses and aretherefore degraded in quality as water flows down-stream. Current government policies do not addressthis issue, and there is no attempt at whole-catchmentmanagement in river basins, which is a fundamentalfeature of water quality protection in western nations.

Finally, the problems related to cost recovery in thesanitation sector could be addressed by the introduc-tion of a single payment for water supply and waste-water disposal, at least for some end-users.This twin-ning of the charges is related to the linking of theproject conception and construction noted above andcould imply that a single authority would be responsi-ble for both elements of the sector. Further studies onthis possibility in Rwanda are recommended.

Environmental Issues in theTelecommunications Sector

Only the following two relatively minor environmen-tal issues are relevant to the telecommunications sectorand neither has particular policy-related implications:• Visual intrusion from land lines extending across

the countryside• The possible presence of polychlorinated naph-

thalenes (PCNs) in telecommunications equipment.The first of these issues is self-explanatory; it is clear

that above-ground land lines servicing the telecommu-nications sector should not be used, at least in areaswith a protected conservation status. Above-groundlines are currently in certain areas of the Akagera Na-tional Park, resulting in high visual intrusion.

PCNs are highly persistent chemicals that havebeen found in telecommunications equipment in cer-tain parts of the world. U.K. authorities have been par-ticularly vigilant regarding PCNs and have classifiedPCNs in telecommunications equipment with the


97

even more toxic polychlorinated biphenyls (PCBs).(See the discussion below on PCBs in the energy sec-tion.) Hence, ascertaining whether PCNs are presentin telecommunications equipment in Rwanda wouldhave merit.

Ecotourism in Rwanda

The current CFR is not intended to address thetourism sector in Rwanda; the CFR is restricted to thefour main vertical sectors (transport, energy, water andsanitation, and telecommunications). Ecotourism,however, is of growing importance in Rwanda, andprotecting the environment is intimately linked toconserving key natural resources. This final note is,therefore, included here, in part because of the rele-vance of PSP to ecotourism.

Rwanda is a beautiful country in general terms, butit also includes three key protected areas of exceptionalquality:• The Virunga National Park, home to about half of

the world’s population of the mountain gorilla• The Akagera National Park, with very significant

habitat diversity and a number of endangered andthreatened species

• The Nyungwe Forest, which has one of the highestconcentrations of primates and monkeys in theworld.Authorities have noted that for ecotourism to be

successful in terms of foreign currency income, at leastthree key tourist resources should be present in anyone country. Each of the above resources in Rwanda istherefore important, not only in its own right, but alsobecause it contributes to generating a critical mass ofsuch sites in Rwanda so that the country can attracttourism from a worldwide audience. While it couldbe argued that the mountain gorillas of the Virungaregion, habituated originally by George Schaller, arethe key resources for ecotourism, the constraints placedon the level of visits to each gorilla group (which arealtogether appropriate and should not be changed)

restrict the total income from such visits to only aboutUS$2.5 million annually. To optimize tourism rev-enues,Rwanda needs to further develop the facilities atVirunga, as well as the other two sites noted, tobroaden the spectrum of tourism options. Facilities arecurrently being developed for tourists at the VirungaNational Park, with golden monkeys currently beinghabituated and with walking trails and kayakingplanned for imminent introduction. By comparison,the tourist facilities available at Akagera are very poor,and those at the Nyungwe Forest are not much better.

Given the present state of the Rwandan economy,these facilities should be developed further, to optimizeforeign income. With the exception of the fees forvisiting the mountain gorillas (which are already quitehigh and should remain so), the charges for entry to theprotected areas can (and should) be raised considerablyonce the tourism facilities have been improved. It iscritical, however, that any further development of thekey conservation areas in Rwanda be completed in anenvironmentally acceptable fashion. Otherwise, eco-tourists will not visit the degraded environments thatwould result. It is also vital that physical encroachmenton the three key protected areas be halted immediatelyand that the levels of conservation be stepped up.Poaching remains a threat in all three areas, although thesituation at the Virunga National Park has been con-trolled admirably in recent years under challenging cir-cumstances. It is also of fundamental importance thatthe local communities become more closely involved inthe conservation efforts and that some of the revenuefrom tourism remain in the local region to improveboth the infrastructure and the economy as a whole.

The use of PSP to hurdle the initial barriers of in-vestment and expertise in the development of eco-tourism facilities has been demonstrated in many partsof Africa and elsewhere in the world. It is clear thatPSP should be a favored option for Rwanda, albeitwithin the constraints of a highly developed monitor-ing and inspection regime designed to conserve thekey natural resources at all costs.

This appendix describes some of the relevant legisla-tion in Rwanda.

Rwandan Land Law

Rwandan land law is a combination of civil law normsand local customary law.The latter is particularly im-portant as far as rural projects such as water and roadprojects are concerned.

Customary Law

The characteristics of customary law are that it is un-written; it gives the right to deal with the product ofthe land (usufruct or superficie); it can be given up by itsowner with prior written consent of the state, givenafter inquiry and indemnity to the owner of the rightsunder customary law; and there is no security oftenure, because the state can confiscate the right at anytime in the public interest. The indemnity referred toabove is again given.

The authority for the application of customary lawis article 98 of the Rwandan constitution, which re-quires that customs meeting the following conditionsbe applied:• The custom has not been modified by existing law.• It does not contradict the constitution or other law

or regulation.• It is not against public order and public morality.

Land Law Reform

The proposed reform of the land law would do the fol-lowing, if it took its current form:• Confirm the powers of the state to confiscate land

for public purposes.• Allow foreigners—who are currently limited to

rights of bail emphytéotique, or concession contracts,of 99 years maximum—to transfer land in the do-maine privé of the state for, among other things, in-dustrial or commercial use. Such transfers would beachieved by ministerial or, in the case of large-scaleprojects, presidential decree.

• Define the land held in public ownership (bothdomaine publique and domaine privé).

• Define land that can be held in private ownership—that is, land held by customary or written rights thatforms neither part of the domaine publique nor thedomaine privé of the state or commune.

• Give the proprietors of customary land rights theability to register them.

• Extend the existing system of land registration.The domaine publique of the state is defined as all

land subject to public use, as well as that forming partof the domaine environnemental of the state, which in-clude the beds of all lakes and rivers and the shores ofall lakes and river banks, as determined by Ministerialdecree, and abreuvoirs (waterholes) and the emprises(banks) of main trunk roads. Surface and groundwaterbelong to no one.

98

Appendix BThe Legal System

Appendix B: The Legal System

99

The domaine privé of the state is composed of allland that does not form part of its domaine publique, ofthe domaine foncier (real estate) of the commune or dis-trict, or of the domaine privé of private persons.

In other words, all ownership of land other thanthat in the domaine publique is a derogation from thebasic principle that land is the property of the state,which has the right (and duty) on behalf of the citizensto monitor its use; to see that where it has been alien-ated it is mise en valeur (properly used); and, if it is notproperly used, to take it back into public ownership.The domaine foncier of the district also comprises a do-maine publique and a domaine privé.

Rights of Ownership

Ownership is defined in the Civil Code as “the right todispose of something in an absolute and exclusivemanner, subject to any restrictions imposed by law andany rights of third parties.” Ownership of land includesthe ownership of the space above and below the land.However if something is done at a height or depth thatwill have no effect on the owner, the owner cannotprevent it.

The owner of the land has no rights over water onthe land, nor over any minerals that may be the subject

of licenses under mining legislation.The bed of everylake and navigable watercourse, whether flottable ounon, forms part of the public property of the state.When a watercourse forms a new bed by abandoningthe old one, the state acquires the new bed but mustcompensate the proprietors of the new land.

The shores of all lakes and banks of rivers andstreams are property of the state for 10 meters from thehighest watermark. If an owner has a water source onthe land that is merely a feeder stream to a watercourse,the owner may use it as he or she wishes. If such asource forms the head of a stream whose bed is distinctfrom neighboring land, the owner may only use it ac-cording to the rules laid down in the Law of 6.5.52.

No one may own the water in streams and lakesand underground watercourses. Subject to the laws orregulations that govern the use of such water and sub-ject to concessions awarded by the public authorities,the right to use such water is common to all.

Clearly, these land and water rights have a consider-able influence on the way in which projects involvingthe private sector can be structured. If reform of theland law is to take place, it should include provisionsmaking it more user friendly to outside investors.

100

Appendix CRoad Maintenance, Rehabilitation,and Reconstruction Costs

Table C.1 Expenditure and Investments Associated with Road Building, Rehabilitation, and Maintenance of VariousTypes of Roads in Rwanda

Approximateexpenditure

intervals Lower cost Upper cost Approximate averageRoad type and action (years) limit (US$/km) limit (US$/km) cost (US$/km)

PavedAnnual maintenance 1 2,300 2,600 2,500Periodic maintenance 5 100,000 120,000 110,000Rehabilitation 450,000 560,000 500,000Reconstruction 560,000 700,000 600,000

Unpaved main roadAnnual maintenance 1 1,600 2,400 40,000Periodic maintenance 5 20,000 30,000 25,000Rehabilitation 30,000 80,000 80,000Reconstruction 80,000 150,000 90,000

Unpaved secondary roadsAnnual maintenance 1 1,500 2,000 1,750Periodic maintenance 5 20,000 30,000 25,000Rehabilitation 30,000 40,000 35,000Reconstruction 40,000 80,000 120,000Sources: Road Maintenance Fund 2002;World Bank 2002; MINITRACO 2002.

Appendix C: Road Maintenance, Rehabilitation, and Reconstruction Costs

101

Table C.2 Expenditure and Investment Required to Rehabilitate the Paved Road Network in Rwanda

Road condition

Paved main roads Good Fair Poor Total

10 years’ percentage 45% 30% 25% 100%Length (kilometers) 418.5 279 232.5 930

Expenditure per 0 250,000 500,000kilometer over 10 yearsof rehabilitation (US$)

Total investment and 0 69,750,000 116,250,000 186,000,000expenditure requiredover 10 years ofrehabilitation (US$)

Total cost over 10 years (US$) 0 69,750,000 116,250,000 186,000,000

Annual cost (US$) 0 6,975,000 11,625,000 18,600,000

Sources: Road Maintenance Fund 2002;World Bank 2002; MINITRACO 2002.

Table C.3 Expenditure and Investment Required to Rehabilitate the Unpaved Main Road Network in Rwanda

Road conditionUnpaved main roads Good Fair Poor Total

10 years’ percentage 10% 40% 50% 100%Length (kilometers) 443.6 1,774.4 2,218 4,436Expenditure per kilometer over 10 years of 0 30,000 90,000

rehabilitation (every 10 years) (US$)

Total investment and expenditure required over 0 53,232,000 199,620,000 252,852,00010 years of rehabilitation (US$)


Annual cost (US$) 0 5,323,200 19,962,000 25,285,200

Sources: Road Maintenance Fund 2002;World Bank 2002; MINITRACO 2002.

Table C.4 Expenditure and Investment Required to Rehabilitate the Unpaved Secondary Road Network in Rwanda

Road condition

Unpaved secondary roads Good Fair Poor Total

10 years’ percentage 10% 40% 50% 100%Length (kilometers) 760 3,040 3,800 7,600Expenditure per kilometer over 0 30,000 80,000

10 years of rehabilitation (every 10 years) (US$)

Total investment and expenditure required over 0 91,200,000 304,000,000 395,200,00010 years of rehabilitation (US$)


Annual cost (US$) 0 9,120,000 30,400,000 39,520,000Sources: Road Maintenance Fund 2002;World Bank 2002; MINITRACO 2002.


102

Table C.6 Investment Projects: Alternative Scenario

Project Rentabilitynumber Project Cost (FRw million) (percent)IR01 Rehabilitation of the Kigali-Kayonza road 10,560 25IR02 Rehabilitation of the Gatuna road 5,480 24IR03 Rehabilitation of the Kayonza-Kagitumba road 3,680 23IR04 Rehabilitation of the Kayonza-Rusumo road 3,020 21TC03 Conservation works for the Butare-Rusizi 2 road 4,860 19TC02 Conservation works for the Ruhengiri-Gisenyi road 2,260 19IR07 Rehabilitation of the Ruhengiri-Cyanika road 810 17TC01 Conservation works for the Kigali-Ruhengiri road 1,860 15IR09 Rehabilitation of the Bugarama-Ruhwa road 210 10

Table C.7 Annual Road Maintenance Projects: Cost Summary

Cost (FRw million)Project number Project 2002 2005 2008

EC01 Annual maintenance of paved roads 1,185 1,296 1,322EC02 Annual maintenance of unpaved classified roads 1,284 1,781 2,346EC00 Total annual maintenance 2,469 3,077 3,668

Table C.5 Investment Projects: Base Scenario

Project Rentabilitynumber Project Cost (FRw million) (percent)IR01 Rehabilitation of the Kigali-Kayonza road 10,560 25IR02 Rehabilitation of the Kigali-Gatuna road 4,470 24IR03 Rehabilitation of the Kayonza-Kagitumba road 3,020 23IR04 Rehabilitation of the Kayonza-Rusumo road 2,470 21IR05 Rehabilitation of the Butare-Cyangugu-Rusizi 2 road 16,410 20IR06 Rehabilitation of the Ruhengiri-Gisenyi road 9,270 18IR07 Rehabilitation of the Ruhengiri-Cyanika road 810 17IR08 Rehabilitation of the Kigali-Ruhengiri road 11,370 12IR09 Rehabilitation of the Bugarama-Ruhwa road 210 10

103

AIPA (Africa Institute for Policy Analysisand Economic Integration). 2002. “TheRwandan Economy: A Strategy forInvestment.” Report prepared for theRwanda Investment Promotion Agency.Kigali.

Austria Rail Engineering (ARE). 1991.“Kagera Basin Railway Study.”

Berocan. 1998.BMI-Techknowledge. 2002. Communica-

tion Technologies Handbook 2002.Johannesburg.

Booz-Allen & Hamilton. 1999. “RwandaDiagnostic Report.”

BUNEP (Bureau National d’Études deProjet). 1986. “Study of the Need for aBoatyard on Lake Kivu.” Kigali.

CIA (Central Intelligence Agency). 2001.The World Factbook. Available online athttp://www.cia.gov/cia/publications/factbook/

References

DECON (Deutsche Energie-ConsultIngenieurgesellschaft mbh). 1998 and1999.

ECA (United Nations Economic Commis-sion for Africa). 1986 (November).“Study for the Development of Trans-port on Lake Kivu.”

ECA (United Nations Economic Commis-sion for Africa). 1986 (December).“Preliminary Study on the Navigabilityof the River Kagera.”

Economist Intelligence Unit. 2000.“National Water Resources Management

Policy.” 1998. Government of Rwanda,Kigali.

Protekem and Common Market of Easternand Southern Africa. 2000. Discussionpaper. Processed.

Scetauroute. 2002. Report on roadsprepared for the European Union.Paris.

Transparency International. 2002. CorruptionPerceptions Index 2002. Berlin.

Water Utility Partnership SBNet. 2001. Re-port of Performance Indicators, African WaterSupply and Sanitation Utilities 2001.Water Utility Partnership, Abidjan, Côted’Ivoire.

World Bank. 2002. “Rwanda Poverty Re-duction Strategy Paper.” Washington,D.C.

World Bank.2003.“Indigenous Investment.”Washington, D.C.

1818 H Street, NW

Washington, DC 20433 USA

Telephone: 202.473.1000

Facsimile: 202.477.6391

Internet: www.worldbank.org


PPIAF Program

Management Unit

c/o The World Bank

1818 H Street, NW, MSN I9-907

Washington, DC 20433 USA

Telephone: 202.458.5588

Facsimile: 202.522.7466

Internet: www.ppiaf.org


PUBLIC-PRIVATE INFRASTRUCTUREADVISORY FACILITY

THE WORLD BANK

ISBN 0-8213-5965-7

Priva

teSo

lutio

ns

for

Infra

structu

rein

Rw

and

a