60803 D I R E C T I O N S I N D E V E LO P M E N T Infrastructure Africa's Transport Infrastructure Mainstreaming Maintenance and Management Ken Gwilliam Africa's Transport Infrastructure Africa's Transport Infrastructure Mainstreaming Maintenance and Management Kenneth Gwilliam with Heinrich Bofinger, Richard Bullock, Robin Carruthers, Ajay Kumar, Mike Mundy, Alberto Nogales, and Kavita Sethi Vivien Foster and Cecilia Briceño-Garmendia, Series Editors © 2011 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org All rights reserved 1 2 3 4 14 13 12 11 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The bound- aries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a 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; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN: 978-0-8213-8456-5 eISBN: 978-0-8213-8605-7 DOI: 10.1596/978-0-8213-8456-5 Library of Congress Cataloging-in-Publication Data Gwilliam, K. M. Africa's transport infrastructure / Kenneth Gwilliam. p. cm. -- (Africa development forum series) Includes bibliographical references and index. ISBN 978-0-8213-8456-5 -- ISBN 978-0-8213-8605-7 (electronic) 1. Transportation--Africa. 2. Infrastructure (Economics)--Africa. I. Title. HE282.A2G87 2011 388.096--dc22 2010047211 Cover photo: Arne Hoel, World Bank Cover design: Debra Naylor, Washington, D.C. Contents About the AICD xix Series Foreword xxi About the Authors xxiii Acknowledgments xxv Abbreviations xxix Chapter 1 The Legacy of History 1 Political History: Colonialism and Independence 1 A Consequence of History: A Distorted Transport Sector 3 The Outcome: High Costs, Poor Service, and Reduced Trade 6 Country Diversity and Uneven Economic Performance 10 A New-Millennium Renaissance 14 Notes 15 References 15 v vi Contents Chapter 2 Roads: The Burden of Maintenance 17 The Road Network 18 Road Infrastructure Performance 29 Institutions: Ongoing Reforms 37 Road Spending: A Problem of Execution 47 Freight Transport: Too Expensive 71 The Way Forward 77 Notes 79 References 80 Chapter 3 Railways: Not Pulling Their Weight 83 Africa's Rail History: Opening Up the Continent 83 A Sparse and Disconnected Network 85 Investment and Maintenance 89 The Market 91 Freight Tariffs: Increasingly Competitive 101 Why Are Railways Uncompetitive? 104 Institutional Arrangements 106 Operational Performance 115 Financial Performance 125 The Way Forward 132 Notes 134 References 137 Chapter 4 Airports and Air Transport: Policies for Growth 139 Airport Infrastructure 139 Operations 152 The Way Forward 176 Notes 177 References 179 Chapter 5 Ports and Shipping: Moving toward Modern Management Structures 181 Coping with Rapidly Changing Trade Patterns 181 The Institutional and Regulatory Framework 195 Infrastructure Development 206 Contents vii Performance, Cost, and Quality 212 The Way Forward 219 Notes 222 References 223 Chapter 6 Urban Transport: Struggling with Growth 225 Infrastructure: Roads 227 Infrastructure: Rails 231 Institutions 232 Services 236 Fares 250 Financing and Subsidies 256 Regulation 260 The Way Forward 265 Notes 267 References 269 Chapter 7 Spending to Improve Connectivity 271 The Expenditure Model in Brief 272 A Detailed Look at the Model's Inputs 275 Applying the Model 284 Outputs of the Model 293 Insights from the Connectivity Analysis 306 Notes 308 References 309 Chapter 8 Financing: Filling the Gaps 311 Expenditures 312 What Can Be Done about the Shortfalls? 322 The Residual Funding Gap 347 The Way Forward 350 Notes 352 References 353 Chapter 9 Governance: The Key to Progress 355 The Context of National Governance 356 Traditions and Attitudes 357 Institutions 363 Capacity 371 The Way Forward 375 viii Contents Notes 377 References 378 Chapter 10 Conclusion: An Agenda for Action 381 Critical Transport Policy Issues 382 Improving Governance 394 Expenditure Requirements 400 Notes 405 References 405 Appendix 1 Introduction 407 Appendix 1a AICD Background Documents Relevant to the Transport Sector 407 Appendix 1b Country Typology for Study Countries 409 Reference 409 Appendix 2 Roads 411 Appendix 2a Road Data Sources and Analysis 411 Appendix 2b Basic Country Data for the Set of 40 Countries 416 Appendix 2c Classified Road Network Length for 40 Countries 418 Appendix 2d Road Network Densities for 40 Countries 420 Appendix 2e Road Network Length by Surface Class and Network Type for 40 Countries 422 Appendix 2f Average Annual Daily Traffic by Road Type for 40 Countries 424 Appendix 2g Distribution of Networks by Traffic Level for 40 Countries 425 Appendix 2h Vehicle Utilization of Roads by Surface Class and by Passenger and Freight for 40 Countries 427 Appendix 2i Classified Road Network Condition by Network Type for 40 Countries 428 Appendix 2j Road Accident Rates for Countries in Africa 430 Appendix 2k Road Maintenance Initiative Institutional Indicators, September 2007 434 Contents ix Appendix 2l Selected Standards by Network Type, Surface Class, and Traffic Level for 40 Countries 437 Appendix 2m Preservation Requirements for Securing the Custom Standard over a 20-Year Period for 40 Countries 439 Appendix 2n Preservation Requirements for Securing the Optimal Standard over a 20-Year Period for 40 Countries 441 Appendix 2o Custom Standard 20-Year Preservation Needs by Work Type for 40 Countries 443 Appendix 2p Optimal Standard 20-Year Preservation Needs by Work Type for 40 Countries 445 References 447 Appendix 3 Rail Transport 449 Appendix 3a Rail Networks in Africa 450 Appendix 3b Production Structure of African Railways, Average 1995­2005 452 Appendix 3c Rail Passenger Traffic 454 Appendix 3d Pricing and Institutions 456 Appendix 3e Factor Productivity 458 Railway Names 460 Reference 461 Appendix 4 Airports and Air Transport 463 Appendix 4a Data Sources for Air Transport Analysis 463 Appendix 4b Airports 466 Appendix 4c City Pairs Served 467 Appendix 4d Installation of Ground-Based Navigational Aids in Africa 469 Appendix 4e Total Capacity Supplied 470 Appendix 4f Costs of Airport Construction versus Rehabilitation 472 Appendix 4g Domestic Air Transport Markets in Africa, 2007 473 Appendix 4h Market Concentration, 2007 474 x Contents Appendix 4i Trends in Aircraft Age 475 Appendix 4j Trends in Aircraft Size 477 Appendix 4k Safety Assessments, 2007 480 Appendix 5 Ports and Shipping 483 Appendix 5a Annual Traffic 484 Appendix 5b Institutional Characteristics 486 Appendix 5c Infrastructure Facilities 488 Appendix 5d Cargo-Handling Performance Indicators 491 Appendix 5e Port Access and Landside Quality 493 Appendix 5f Average Port Costs and Charges 496 Appendix 6 Expenditure Needs 499 Appendix 6 Costs of Achieving Targets of Pragmatic Scenario, by Expenditure Purpose 499 Appendix 7 Financing 501 Appendix 7a Transport Spending and Finance Sources, by Country 502 Appendix 7b Potential Efficiency Gains 504 Appendix 7c Agency and Total Social Benefits of Timely Road Maintenance 506 Appendix 7d Closing the Gap 513 Note 515 Index 517 Boxes 1.1 The Economic Costs of Political Fragmentation: The Case of Guinea 5 2.1 Road Concessions in Africa 50 4.1 Air Afrique 156 5.1 Inland Waterways: A Neglected Asset 186 5.2 The Nigeria Port Concessions 203 6.1 Introducing Dedicated Infrastructure for Bus Transit 235 6.2 Financing Large vs. Small Buses in Nairobi 258 7.1 Mixing Scenarios 299 9.1 The Kenyan Government's Purchase of Luxury Vehicles for Official Use 358 Contents xi Figures 2.1 Range of Primary and Secondary Road Densities 21 2.2 Range of Tertiary and Unclassified Road Densities 23 2.3 Spatial Density of Road Networks in World Regions 24 2.4 Total Road Networks per Capita in World Regions 24 2.5 Total Road Network as Share of GDP in World Regions 25 2.6 Volume of Traffic Carried on Main Network 27 2.7 Volume of Traffic Carried on Rural Network 28 2.8 Distribution of Road Network Length across Condition Classes 30 2.9 Road Asset Value as Percentage of the Potential Maximum 34 2.10 Correlation between Percentage of Main and Rural Roads in Good Condition 35 2.11 Estimated RAI from Current Network and Percentage of Expansion Needed to Reach 100 Percent RAI 35 2.12 Accessibility and Agricultural Production 36 2.13 Fuel Levy Relative to Optimal Requirements for Maintenance and Rehabilitation 39 2.14 Average Fuel Levy across Countries with Second-Generation Road Funds, 2007 40 2.15 Overview of Road Fund Allocation Rules 42 2.16 Evaluation of Road Fund Reforms 43 2.17 Average Annual Expenditures on Road Transport by Country, 2001­05 48 2.18 Percentage of Road Spending Allocated to Capital Projects 51 2.19 Capital Budget Execution Ratios 53 2.20 Deviation of Capital Expenditure from Expenditure Required to Meet Rehabilitation Requirements within a Five-Year Period 54 2.21 Foreign Funding as Percentage of Capital Spending 56 2.22 Relationship between Capital Spending and Maintenance Spending per Kilometer of Main Network 57 2.23 Average Maintenance Spending across Different Parts of the Network 58 2.24 Deviation of Actual Maintenance Expenditure from That Required to Attain Custom Standard of Maintenance 59 xii Contents 2.25 Aggregate Requirements over a 20-Year Period for Preserving the Road Networks as Percentage of the Current Annual GDP 61 2.26 Extent of Over- and Underengineering on Main Road Networks 64 2.27 Extent of Over- and Underengineering on Rural Road Networks 65 2.28 Road Asset Value as Percentage of GDP 68 2.29 Relationship between Road Networks in Poor Condition and the Climate-Terrain Index 70 2.30 Relationship between Road Networks in Good Condition and Their Score on the Road Fund Quality Index 72 2.31 Relationship between Road Networks in Good Condition and Maintenance Expenditures 73 3.1 Passenger and Freight Traffic of Railways in Sub-Saharan Africa (Annual Average, 1995­2005, Excluding Spoornet) 93 3.2 Annual Average Distance Traveled on Railways in Sub-Saharan Africa, 1995­2005 94 3.3 Traffic Growth on Railways in Sub-Saharan Africa 96 3.4 Comparison of Bus and Rail Fares and Travel Times 99 3.5 Commodities Carried on Select Railways in Sub-Saharan Africa 100 3.6 Average Tariffs for Passenger and Freight Traffic, 1995­2005 102 3.7 Average Tariffs by Commodity for Select Railways in Sub-Saharan Africa 103 3.8 Indicative Freight Rates: 12-meter Container Inland from Port, 2003 105 3.9 Labor Productivity of Railways in Sub-Saharan Africa 118 3.10 Passenger Car Productivity 120 3.11 Freight Wagon Productivity 121 3.12 Labor and Asset Productivity for Four Railways before and after Concessioning 123 3.13 Cost per Traffic Unit for Railways in Sub-Saharan Africa 126 3.14 Passenger Service Cost Recovery, 2002 128 3.15 Freight Cost Recovery, 2002 129 3.16 Financing Structure of Select Concessions 130 3.17 Concession Fees 131 4.1 Airport Charges Overall, by Aircraft Type across 19 Sample Airports 146 Contents xiii 4.2 Estimated International Passenger Capacity between 2001 and 2007, as Measured in Seat-Kilometers 155 4.3 Trends in Aircraft Age, 2001­07 163 4.4 Trends in Aircraft Type, 2001­07 164 4.5 Pricing of Flights within Africa versus Intercontinental Flights 168 4.6 ICAO Analysis of Safety Implementation 175 5.1 Container Trade Development Summary in the Study Ports, by Region, 1995 and 2005 182 5.2 Top 10 Sub-Saharan African Ports 183 5.3 The Imbalance of Sub-Saharan African Container Trade, 2005 189 5.4 Development of General Cargo Traffic, 1995­2005 191 5.5 Oil Production and Proven Reserves in Sub-Saharan Africa 193 5.6 Evolution of Private Sector Participation in African Ports 201 5.7 Container-Handling Systems at Major Ports of the Region 213 5.8 Container-Handling Performance by Equipment Type and Port Management Type 215 5.9 Typical Truck Cycle and Dwell Times by Region 216 5.10 Performance in General Cargo Handling in Individual Ports of Sub-Saharan Africa 217 5.11 Typical Gateway Container-Handling and General Cargo Charges Applied in Major African and World Ports 220 6.1 Condition of Urban Roads in 20 AICD Countries 229 7.1 A Model for Estimating the Costs of Transport Infrastructure Expenditure Needs 273 7.2 Scenario Costs in Countries Where the Base Scenario Would Cost More than 5 Percent of GDP 298 7.3 Comparison of Scenarios by Type of Spending 303 7.4 Comparison of Scenarios by Mode of Transport 304 7.5 Rural Connectivity Costs for Different Percentages of Agricultural Production, by Value 306 7.6 Rural Connectivity Costs for Low-Income, Fragile States with Different Categories of Roads 307 8.1 Sources of Finance for Transport Spending in Africa by Country 316 8.2 Sources of Funding for Transport Infrastructure Capital Investment 317 xiv Contents 8.3 Capital Investment and O&M Spending for Transport from All Sources by Country 318 8.4 Transport Spending as Percentage of Needs 319 8.5 Needs and Spending as Percentage of GDP 321 8.6 Transport Infrastructure Assets in Need of Rehabilitation 325 8.7 Efficiency Gains from Improved Maintenance 328 8.8 Potential for Reallocation of Spending 330 8.9 Cost of Redressing Infrastructure Backlogs over Varying Time Frames 332 8.10 Potential Efficiency Gains from Different Sources 334 8.11 Overview of Private Commitments to African Transport Infrastructure 341 8.12 Costs of Capital by Funding Source 346 8.13 Transport Infrastructure Financing Gap, Assuming All Efficiency Gains 348 8.14 Transport Infrastructure Financing Gap, Excluding Any Efficiency Gains 349 A2a.1 Matrix of Road Classes: Overall Network Evaluation 413 Maps 1.1 World Logistics Performance Index 7 1.2 A Country Typography of Sub-Saharan Africa 10 3.1 The African Rail Network in 2009 85 3.2 Railway Concessions Awarded in Africa since 1990 108 4.1 International Connectivity in Sub-Saharan Africa: Winners and Losers 2001­07 140 4.2 Top 30 Intercontinental Routes for Sub-Saharan Africa as of November 2007 154 4.3 Regional Growth Zones in Seats Offered 157 4.4 African Countries Potentially Served by Commuter-Style Turboprop Aircraft Using a Hub in Lagos 167 Tables 1.1 Typology of Countries 12 1.2 Basic Characteristics of African Countries vs. World's Other Developing Countries 13 2.1 Overview of Africa's Key Transport Corridors for International Trade 19 2.2 Cross-Regional Comparison of Paved Road Infrastructure in Low-Income Countries 26 Contents xv 2.3 Trends in Road Condition, 2004­07 32 2.4 Average Annual Expenditures per Kilometer of Main Road by Country Category, 2001­05 51 2.5 Percentage of Road Spending Allocated to Capital Projects, by Country Category 52 2.6 Capital Budget Execution Ratios, by Country Category 53 2.7 Capital Expenditure as Percentage of Rehabilitation Needs, by Country Category 55 2.8 Actual Maintenance Expenditure as Percentage of Expenditure Required for Custom Maintenance Standard, by Country Category 60 2.9 Recent Estimates of Unit Costs of Road Maintenance and Rehabilitation 62 2.10 Performance of the International Gateway Corridors 74 3.1 Key Features of Concessions, 1993­2008 110 3.2 Initial Concession Shareholdings 116 4.1 Airport Terminal Capacity vs. Reported Passengers and Estimated Seats 144 4.2 Overall Runway Quality in Africa 145 4.3 Public-Private Investments in African Airports 148 4.4 Estimated Seats and Growth Rates in African Air Transport Markets 152 4.5 Top 14 Airports in Africa Serving International Travel within Africa 157 4.6 Top 15 Airlines Overall in the African Passenger Market 159 4.7 Top 15 Airlines Providing International Service within Africa 160 4.8 Breakdown of Aircraft Age for Analysis 162 4.9 Breakdown of Aircraft Type by Market Size 165 4.10 Air Service Liberalization among Regional Groupings in Africa 170 4.11 Percentage of Flights between Country Pairs Served by Airlines Not Based in Either Country 173 5.1 Regional Ports and Their Transit Traffic Markets--Actual and Potential 185 5.2 Major African Ports and Their Transshipment Potential 187 5.3 Dry Bulk Operations in Africa 192 5.4 Port Management Models by Region 200 5.5 Private Sector Transactions for All Port Sectors in Africa 202 xvi Contents 5.6 Major Institutional Reform Initiatives Implemented or Planned for African Ports, as of 2009 204 5.7 National Institutions Responsible for Port Regulation in Africa, by Type 206 5.8 Annual Port Capacity and Current Demand, Selected Ports 207 5.9 Principal New Port Developments in Africa 209 5.10 Gantry Crane Productivity, 2004 214 5.11 Performance in General Cargo Handling across Regions of Africa 218 6.1 Size and Other Characteristics of 14 African Cities 226 6.2 Characteristics of City Road Networks 228 6.3 Institutions with Responsibility for Public Transport in Cities 233 6.4 Modal Shares of Transport in Cities 237 6.5 Large-Bus Operations in Cities 240 6.6 Characteristics of Minibus Services in Cities 242 6.7 Average Bus Age and Fleet Size 244 6.8 Average Distance Traveled by Large Buses and Minibuses 245 6.9 Availability of Public Transportation in Cities 246 6.10 Fuel Prices in Cities, July 2007 249 6.11 Fare-Setting Procedures for Large Buses and Minibuses 252 6.12 Average Bus Fare 253 6.13 Spending on Urban Transport as a Share of Household Income 255 6.14 Regulatory Framework for Urban Public Transport 261 7.1 Definition of the Base Scenario 282 7.2 Definition of the Pragmatic Scenario 283 7.3 Transport Infrastructure: Current, Base Scenario, and Pragmatic Scenario 286 7.4 Regional Airports: Current Characteristics vs. Characteristics Needed to Meet Connectivity Target 287 7.5 Road Lengths Needed to Reach Rural Connectivity Standards 289 7.6 Sources of Road Upgrades to Achieve Rural Connectivity Targets 290 7.7 Types of Road Counted toward Urban Connectivity Targets 290 7.8 Average Annual Investment Needs: Base Scenario, 2006­15 295 Contents xvii 7.9 Average Annual Investment Needs: Pragmatic Scenario 297 B7.1 Comparing Blends of the Base and Pragmatic Scenarios 299 7.10 Transport Infrastructure Expenditure as Percentage of GDP, by Country Group and Scenario 300 7.11 Highest Transport Infrastructure Spending Needs by Country, Relative to GDP, Population, and Land Area 302 7.12 Spending Needs by Purpose for Each Country Group 304 7.13 Percentage Allocation of Investment Needs by Country Group and Transport Mode (base scenario) 305 8.1 Annual Transport Spending by Finance Source and Country Type 313 8.2 Transport Spending as a Share of Total Budget in Africa 315 8.3 Average Percentage of Capital Budget Actually Spent, by Country Type 324 8.4 Overall Value of Agency and Vehicle-Operating Cost Savings 329 8.5 Economic Rates of Return for Key Infrastructure Interventions in Africa 331 8.6 Potential Gains from Improved Efficiency in Transport Spending 333 8.7 Net Change in Central Government Budgets by Economic Use, 1995­2004 336 8.8 Outstanding Financing Stock for Transport Infrastructure as of 2006 344 10.1 Sectorwide Policy Requirements: An Action Program 386 10.2 Roads and Road Transport: An Action Program 388 10.3 Rail Transport: An Action Program 390 10.4 Airports and Air Transport: An Action Program 392 10.5 Ports and Maritime Transport: An Action Program 393 10.6 Urban Transport: An Action Program 395 10.7 Overcoming Institutional Weaknesses: An Action Program 398 10.8 Overcoming Behavioral Weaknesses: An Action Program 401 A1a.1 Background Papers 408 A1a.2 Working Papers 408 A2a.1 RONET Default Assignment of Traffic Levels 414 xviii Contents A2j.1 Reported and Estimated Deaths from Road Accidents, 2008 431 A7c.1 Unit Cost of Road Works 507 A7c.2 Comparison of Road Agency Costs for Various Roads and for Rehabilitation vs. Periodic Maintenance 508 A7c.3 Total Economic Costs Discounted at 12 Percent 510 About the AICD This study is a product of the Africa Infrastructure Country Diagnostic (AICD), a project designed to expand the world's knowledge of physical infrastruc- ture in Africa. The AICD provides a baseline against which future improvements in infrastructure services can be measured, making it possible to monitor the results achieved from donor support. It also offers a more solid empirical foundation for prioritizing invest- ments and designing policy reforms in the infrastructure sectors in Africa. The AICD was based on an unprecedented effort to collect detailed economic and technical data on the infrastructure sectors in Africa. The project produced a series of original reports on public expenditure, spend- ing needs, and sector performance in each of the main infrastructure sectors, including energy, information and communication technologies, irrigation, transport, and water and sanitation. The most significant findings were synthesized in a flagship report titled Africa's Infrastructure: A Time for Transformation. All the under- lying data and models are available to the public through a Web portal (http://www.infrastructureafrica .org), allowing users to download customized data reports and perform various simulation exercises. The AICD was commissioned by the Infrastructure Consortium for Africa following the 2005 G-8 Summit at Gleneagles, which flagged the importance of scaling up donor finance to infrastructure in support of Africa's development. The first phase of the AICD focused on 24 coun- tries that together account for 85 percent of the gross domestic product, population, and infrastruc- ture aid flows of Sub-Saharan Africa. The countries were Benin, Burkina Faso, Cape Verde, Cameroon, Chad, Democratic Republic of Congo, Côte d'Ivoire, xix xx About the AICD Ethiopia, Ghana, Kenya, Lesotho, Madagascar, Malawi, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, South Africa, Sudan, Tanzania, Uganda, and Zambia. Under a second phase of the project, coverage was expanded to include the remaining countries on the African continent. Consistent with the genesis of the project, the main focus was on the 48 countries south of the Sahara that face the most severe infrastructure chal- lenges. Some components of the study also covered North African countries to provide a broader point of reference. Unless otherwise stated, therefore, the term "Africa" is used throughout this report as a shorthand for "Sub-Saharan Africa." The AICD was implemented by the World Bank on behalf of a steering committee that represents the African Union, the New Partnership for Africa's Development (NEPAD), Africa's regional eco- nomic communities, the African Development Bank, and major infrastructure donors. Financing for the AICD was provided by a multidonor trust fund to which the main contributors were the Department for International Development (United Kingdom), the Public Private Infrastructure Advisory Facility, Agence Française de Développement, the European Commission, and Germany's Kreditanstalt für Wiederaufbau (KfW). The Sub-Saharan Africa Transport Policy Program and the Water and Sanitation Program provided technical support on data collection and analysis pertaining to their respec- tive sectors. A group of distinguished peer reviewers from policy-making and academic circles in Africa and beyond reviewed all of the major outputs of the study to ensure the technical quality of the work. Following the completion of the AICD project, long- term responsibility for ongoing collection and analysis of African infrastructure statistics was transferred to the African Development Bank under the Africa Infrastructure Knowledge Program (AIKP). A second wave of data collection of the infrastructure indicators analyzed in this volume was initiated in 2011. Series Foreword The Africa Infrastructure Country Diagnostic (AICD) has produced continent-wide analysis of many aspects of Africa's infrastructure chal- lenge. The main findings were synthesized in a flagship report titled Africa's Infrastructure: A Time for Transformation, published in November 2009. Meant for policy makers, that report necessarily focused on the high-level conclusions. It attracted widespread media coverage feeding directly into discussions at the 2009 African Union Commission Heads of State Summit on Infrastructure. Although the flagship report served a valuable role in highlighting the main findings of the project, it could not do full justice to the richness of the data collected and technical analysis undertaken. There was clearly a need to make this more detailed material available to a wider audience of infrastructure practitioners. Hence the idea of producing four technical monographs, such as this one, to provide detailed results on each of the major infrastructure sectors--information and communication technologies (ICT), power, transport, and water--as companions to the flagship report. These technical volumes are intended as reference books on each of the infrastructure sectors. They cover all aspects of the AICD project relevant to each sector, including sector performance, gaps in financing and efficiency, and estimates of the need for additional spending on xxi xxii Series Foreword investment, operations, and maintenance. Each volume also comes with a detailed data appendix--providing easy access to all the relevant infrastructure indicators at the country level--which is a resource in and of itself. In addition to these sector volumes, the AICD has produced a series of country reports that weave together all the findings relevant to one par- ticular country to provide an integral picture of the infrastructure situa- tion at the national level. Yet another set of reports provides an overall picture of the state of regional integration of infrastructure networks for each of the major regional economic communities of Sub-Saharan Africa. All of these papers are available through the project web portal, http://www.infrastructureafrica.org, or through the World Bank's Policy Research Working Paper series. With the completion of this full range of analytical products, we hope to place the findings of the AICD effort at the fingertips of all interested policy makers, development partners, and infrastructure practitioners. Vivien Foster and Cecilia Briceño-Garmendia About the Authors The lead author, Ken Gwilliam, retired from the post of economic adviser to the transport sector of the World Bank in 2002. During a decade in that role, he was the author of Bank policy documents and papers on transport strategy (Sustainable Transport, 1996), urban transport (Cities on the Move, 2002) and transport and air pollution (Air Pollution from Mobile Sources, 2004). Prior to this, he was professor of transport economics and director of university transport research institutes in the United Kingdom (University of Leeds) and the Netherlands (Erasmus University Rotterdam). He was for six years a director of the British National Bus Company, editor of the Journal of Transport Economics and Policy, and a member of numerous British and European committees and commissions. Henrich Bofinger is an air transport specialist consulting with the World Bank in Africa. Dick Bullock is a rail specialist consulting with the World Bank in Africa. Robin Carruthers, now retired, was a Lead Transport Economist at the World Bank with a special focus on multimodal sectorwide planning. Ajay Kumar is a Lead Transport Economist with the Africa Region of the World Bank and has been leading the practice on urban transportation issues. xxiii xxiv About the Authors Mike Mundy is a ports specialist with Ocean Shipping Consultants work- ing across Africa. Alberto Nogales is a roads specialist consulting with the World Bank in Africa. Kavita Sethi is a senior transport economist with the Africa Region of the World Bank. Acknowledgments This book was authored by Kenneth Gwilliam, under the overall guid- ance of series editors Vivien Foster and Cecilia Briceño-Garmendia. All the Africa Infrastructure Country Diagnostic (AICD) transport work was undertaken in the framework of a partnership with the Sub-Saharan Africa Transport Policy Program. The book draws upon a number of background papers that were pre- pared by World Bank staff and consultants, under the auspices of the AICD. Excluding the chief author, key contributors to the book on a chapter-by-chapter basis were as follows. Chapter 2 Contributors Rodrigo Archondo Callao, Kavita Sethi, Vivien Foster, Alberto Nogales, Supee Teravaninthorn, Gael Raballand Key AICD Source Documents "The Burden of Maintenance: Roads in Sub-Saharan Africa," Background Paper 14, AICD "Transport Prices and Costs in Africa: A Review of the Main International Corridors," Working Paper 14, AICD xxv xxvi Acknowledgments Chapter 3 Contributors Richard Bullock, Mapapa Mbangala, Pierre Pozzo di Borgo, Lucien Andre Aegerter Key AICD Source Document "Railways in Sub-Saharan Africa," Background Paper 11, AICD Chapter 4 Contributors Heinrich Bofinger, Charles Schlumberger Key AICD Source Document "Challenges to Growth in Africa's Air Transport Industry," Background Paper 16, AICD Chapter 5 Contributors Mike Mundy, Andrew Penfold, Bradley Julian, Michel Luc Donner, C. Bert Kruk Key AICD Source Document "Beyond the Bottlenecks: Ports in Sub-Saharan Africa," Background Paper 8, AICD Chapter 6 Contributors Ajay Kumar, Fanny Barrett Key AICD Source Document "Stuck in Traffic: Urban Transport in Africa," Background Paper 1, AICD Chapter 7 Contributors Robin Carruthers, Ranga Rajan Krishnamani, and Siobhan Murray Key AICD Source Document "Improving Connectivity: Investing in Transport Infrastructure in Sub-Saharan Africa," Background Paper 7, AICD Acknowledgments xxvii Chapter 8 Contributors Cecilia Briceño-Garmendia, William Butterfield, Chuan Chen, Jacqueline Irving, Astrid Manroth, Nataliya Pushak, Afua Sarkodie, Karlis Smits Key AICD Source Document "Financing Public Infrastructure in Sub-Saharan Africa: Patterns, Issues, and Options," Background Paper 15, AICD Chapter 9 Contributors Bruce Thompson, Kavita Sethi Key AICD Source Documents The largest collection of primary data was in the roads sector where two consulting firms carried out the AICD data collection in the field for 40 countries: · Africon in association with InfraAfrica, led by Paul Lombard, covered mostly anglophone countries, and · Tecsult Inc., led by Denis Baron, collected data from most francopho- ne countries. Among the many World Bank staff who helped set up arrange- ments for data collection, we would particularly like to thank Alain Labeau, Anil Bhandari, and Supee Teravaninthorn, AFTTR Coordinators; and Pauline de Curieres de Castelnau, AFTTR Program Assistant, all based in Washington, DC; Alexandre Dossou, Senior Transport Specialist based in the Democratic Republic of Congo; Aguiratou Savadogo-Tinto, Senior Transport Specialist based in Burkina Faso; Ibou Diouf, Senior Transport Specialist based in Côte d'Ivoire; John Kobina Richardson, Transport Specialist based in Ghana; Kingson Khan Apara, Senior Transport Specialist based in Gabon; and Lavite Victorio Ocaya, Senior Highway Engineer based in Uganda. None of the research would have been possible without the generous collaboration of government officials in the key sector institutions of each country, as well as the arduous work of local consultants who assembled the information in a standardized format. xxviii Acknowledgments The data were analyzed using the ROad Network Evaluation Tools (RONET) model designed, developed, and adjusted to satisfy the spe- cific needs of the AICD study by Rodrigo Archondo, Senior Highway Engineer, World Bank. We would particularly like to thank David Luyimbazi (Uganda), Godwin Brocke (Ghana), Atanásio Mugunhe (Mozambique), and Joseph Lwiza (Tanzania), from the road agencies that applied the RONET model in the four initial countries where the model was tested. Seabury ADG gave access to the basic data on which much of the air transport service analysis is based. The AICD also used and adapted Institutional Data collected by the Sub-Saharan Africa Transport Policy Program (SSATP), as well as many SSATP publications referred to in the text. Mustapha Benmmaamar, Senior Transport Specialist, is particularly thanked for his contribution in this respect. The book benefited from widespread peer review by colleagues within the World Bank, notably Michel Donner, C. Bert Kruk, Pierre Pozzo di Borgo, and Kavita Sethi. The external peer reviewer for this volume, Bruce Thompson, provided constructive and thoughtful comments. Alberto Nogales, Carolina Dominguez, and Nataliya Pushak prepared statistical and graphic materials both for the text and the appendixes. The compre- hensive editorial effort of Steven Kennedy is much appreciated. Abbreviations AADT annual average daily traffic AC asphalt concrete ACCO Association des Chauffeurs du Congo ACI Airports Council International ACSA Airports Company South Africa ADC Aeroports du Cameroun ADG Airline Data Group ADS-B automatic dependent surveillance-broadcasts AfDB African Development Bank AGETU Agence de Gestion des Transports Urbains (urban transport agency) AICD Africa Infrastructure Country Diagnostic AMU Arab Maghreb Union ANS air navigation services ASECNA Agence pour la Sécurité de la Navigation Aérienne en Afrique et à Madagascar ATC air traffic control ATRACO Association pour le Transport Commun BAG Banjul Accord Group BR Botswana Railway xxix xxx Abbreviations BRT bus rapid transit c/ntkm cents/net tonne-kilometer c/pkm cents/passenger-kilometer CAA civil aviation authority CAPEX capital expenditure CCFB Companhia dos Caminhos de Ferro da Beira (Beira Railway Company) CDN Corredor de Desenvolvimento do Norte CEAR Central East African Railways Company CEMAC Economic and Monetary Community of Central Africa CETUD Conseil Exécutif des Transports Urbains de Dakar (Executive Council of Dakar Urban Transport) CFCO Chemin de fer Congo-Océan CFM Caminhos de Ferro de Moçambique CFMK Chemins de Fer Matadi Kinshasa CICOS Commission Internationale du Bassin Congo- Oubangui-Sangha COMESA Common Market for Eastern and Southern Africa COSCAPS Cooperative Development of Operational Safety and Continuing Airworthiness Programme DARCOBOA Dar es Salaam Commuter Buses Owners Association DDD Dakar Dem Dikk DRCTU Direction de la Régulation et du Contrôle du Transport Urbain (Directorate of Traffic Regulation and Urban Transport) EAC East African Community ECOWAS Economic Community of West African States EDIFACT Electronic Data Interchange for Administration, Commerce, and Transport EU European Union FAA Federal Aviation Administration (United States) FCE Fianarantsoa Côte Est Railway GDP gross domestic product GIS geographic information system GPRTU Ghana Private Road Transport Union GPS global positioning system HDM-4 Highway Development and Management Model IATA International Air Transport Association IASA International Aviation Safety Assessment ICAO International Civil Aviation Organization Abbreviations xxxi ICTSI International Container Terminal Services Inc. IDA International Development Association IFI international financial institution ILS instrumented landing system IMF International Monetary Fund IMO International Maritime Organization IOSA IATA Operational Safety Audit Ircon Indian Railways Construction Corporation ISPS International Ship and Port Facility Security ISSG Industry Safety Strategy Group IT information technology KBS Kenya Bus Service km kilometer(s) km2 square kilometer(s) km/hr kilometers per hour KRC Kenya Railways Corporation LAMATA Lagos Metropolitan Area Transport Authority LIC low-income country LPI logistics performance index m2 square meter(s) MIC middle-income country mm millimeter(s) MMT Metro Mass Transit Ltd. (Ghana) MVOA Matatu Vehicle Owners Association NPV net present value NRSC national road safety council ntkm net tonne-kilometer NURTW National Union of Road Transport Workers OCBN Organisation Commune Benin-Niger des Chemins de Fer et des Transports (Benin-Niger Railway and Transport Organization) ODA official development assistance OECD Organisation for Economic Co-operation and Development O&M operations and maintenance ONATRACOM Office National de Transport en Commun OPEX operating expenditure PIARC World Road Association pkm passenger-kilometer PMU project management unit xxxii Abbreviations PPI Private Participation in Infrastructure PPIAF Public-Private Infrastructure Advisory Facility PPP public-private partnership PROTOA Progressive Transport Owners Association PSO public service obligation PV present value RAI Rural Accessibility Index REC regional economic community ROCKS Road Costs Knowledge System RONET Road Network Evaluation Tool rpkm revenue passenger-kilometer RSZ Railway Systems of Zambia RVRC Rift Valley Railways Consortium SAA South African Airways SACU Southern African Customs Union SADC Southern African Development Community SARCC South Africa Rail Commuter Corporation SCCF Société Camerounaise des Chemins de Fer SEFICS Société d'Exploitation Ferroviaire des Industries Chimiques du Sénégal SETRAG Société Transgabonnaise SICTA Société Ivoirienne de Contrôle Techniques Automobiles et Industriels SNCC Societe Nationale de Chemins de Fer Congolais (Congolese National Railways) SNTMVCI Syndicat National des Transportent de Marchandes et Voyageurs de Côte d'Ivoire SOCATUR Société Camerounaise de Transports Urbains SOFIB Société Ferroviaire Ivoiro-Burkinabè SOLAS Safety of Life at Sea Convention SOTRA Société de Transport Abidjanais SOTRACO Société de Transport en Commun SOTUC Societé de Transports Urbains du Cameroun SSATP Sub-Saharan Africa Transport Policy Program ST surface treatment STUC Société des Transports Urbains du Congo SUMATRA Surface and Marine Transport Regulatory Authority TEU twenty-foot equivalent unit TRC Tanzania Railways Corporation TU traffic unit Abbreviations xxxiii UNCTAD United Nations Conference on Trade and Development UNECA United Nations Economic Commission for Africa UPETCA Union des Propriétaires de Taxis Compteurs d'Abidjan URC Uganda Railways Corporation UTODA Uganda Taxi Operators and Drivers Association WACEM West African Cement Company WAEMU West African Economic and Monetary Union WBI World Bank Institute WGI Worldwide Governance Indicators YD Yamoussoukro Decision ZR Zambia Rail CHAPTER 1 The Legacy of History This book is about transport in Africa, where Africa is defined to exclude the six countries and one disputed territory generally called North Africa (Algeria, the Arab Republic of Egypt, Libya, Morocco, Sudan, Tunisia, and Western Sahara). What is often referred to as Sub-Saharan Africa will be referred to here as Africa. The main purpose of this book is to assess the factors that affect the performance of Africa's transport infrastructure. While the book is not about geography or political history, a brief review of the fortuities of natural resource endowment and the vicissi- tudes of history is necessary to understand the current state of transport infrastructure and the distortions of transport operations. Political History: Colonialism and Independence Africa's rich natural endowment of diamonds, gold, and other mineral deposits was the attraction that eventually led the industrial powers of Europe to colonize the continent. Later, oil became an even more valu- able prize. Though these resources are concentrated in a broad band of states in Central and West Africa, other countries not formally classified as resource rich, such as South Africa, also have substantial mineral resources. In many parts of the continent, agricultural products such as 1 2 Africa's Transport Infrastructure rubber, coffee, cocoa, and cotton also have high export potential. While the colonial powers exploited Africa's rich resources, most of its popula- tion remained dependent on subsistence agriculture. By 1945, the whole of the continent--with the exception of Ethiopia and Liberia--had been colonized by one European state or another, with the Union of South Africa gaining independence in 1910. Belgium, France, Great Britain, Italy, Portugal, and, in a smaller way, Spain, all had a stake in Africa--as Germany had until the end of World War I. While World War II did not change the basic face of colonial Africa, it renewed European powers' interest in developing their colonies as sources of materials for the war effort. During this period, many transport facilities were built, primarily for the exploitation and export of natural resources. More recently, China's investment in railways has been motivated by that country's need to secure supplies of scarce minerals critical to its growth. The European powers adopted radically different approaches to the political structures of their colonies. Belgium and Portugal did not permit any political activity at all in their territories. Great Britain governed each of its 14 territories separately, allowing a degree of self-determination on internal matters in some. France viewed the African colonies as an integral and indissoluble part of metropolitan France, with entirely parallel politi- cal systems and processes. Whatever the system of administration, pressure for independence grew rapidly in the post­World War II years until, even- tually, the winds of change swept across the colonial territories. The Belgian Congo gained independence in 1960, and by the end of 1968, all the British and French colonies were independent. Portugal withdrew from Angola and Mozambique by 1975. In many cases, the leaders of the independence movements became heads of the newly formed states. Precolonial African societies have been described by Meredith (2005, 154) as "a mosaic of lineage groups, clans, villages, chiefdoms, kingdoms and empires with shifting and indeterminate frontiers and loose alle- giances." He argues that colonial administrators actually oversimplified and hence accentuated ethnic distinctions in their zeal to classify indigenous populations for administrative purposes. The countries that emerged at independence were to a large extent the artificial constructs of colonialism, through which tribal divisions became more deeply entrenched. Many subsequent civil conflicts, such as that in Rwanda, were in part the result of this emphasis on the identification and manip- ulation of tribal groupings. In the early days of independence, many had expected that the interests of nation building would supersede ethnic divisions and lead to greater union. Instead, ethnic divisions increasingly The Legacy of History 3 dominated and fractured the political processes of many countries, including Ghana, Nigeria, and, saddest of all, Rwanda. The independence process itself had significant effects on the political shape of Africa. After France's initial expulsion from Guinea, to which it reacted by withdrawing all resources and support, France shifted to a policy of restructuring before liberating its African colonies. Both French West Africa and French Equatorial Africa were split into multi- ple independent countries with the intent of maintaining French inter- est and influence on the continent. Given the historical association of the colonies with metropolitan France, and given the experience that a number of African politicians had obtained in French government, this strategy did in fact perpetuate strong French influence in a number of countries, such as Côte d'Ivoire. The initial stages of each country's independence were critical. The most significant consequence of the new order was the emergence in many countries of a one-party system. While this system was initially defended as appropriate for nation building in states with multiple ethnic communities, many of the liberators became dictators. Many new national leaders adopted the theories of Marx and Lenin, though not all interpreted socialism the same way. In Ghana, President Kwame Nkrumah saw the path of development in terms of rapid indus- trialization; in Tanzania, President Julius Nyerere saw it in terms of agri- cultural self-sufficiency. Most new leaders, however, shared the belief that the state should direct economic activity, implying strict government con- trol, if not full ownership, of most productive sectors. A Consequence of History: A Distorted Transport Sector This political history has had profound economic consequences for the transport sector, bequeathing a legacy of structural and institutional distortions from which it has still not completely escaped. Several component elements of distortion can be identified. Networks were incomplete. Colonialism was about the exploitation of natural resources. Colonial government administration was typically set- tled in a capital city, often a port, and had little concern for inland passen- ger transport. The infrastructure it developed was usually limited to whatever was deemed necessary for the export of minerals or agricultural products. Only the links between the port and the material source (which might be in one of the neighboring landlocked countries) were of prime interest. The result was that transport networks were extensive in linking 4 Africa's Transport Infrastructure ports and distant sources rather than intensive in giving good network coverage to the whole of the territory. Rail development was emphasized. For heavy, bulk movements over long distances, rail transport was usually believed to have a comparative advan- tage over other modes--particularly road transport. So major investments were made in ports and rail systems. Moreover, because speed was not essential, the rail systems were built to modest technical specifications, with the consequence that once roads began to be developed, the railways were not well equipped to compete in the more time-sensitive passenger transport markets. Systems were distorted by national fragmentation. Independence was accompanied by national fragmentation--a deliberate policy in the case of the former French colonies. Such radical political subdivision of already small postcolonial economies took a heavy toll on the welfare of African citizens (Collier and Venables 2008). In the private sector, subdivision frustrated scale economies and skewed the structure of the overall econ- omy toward peasant agriculture. In the public sector, the small scale raised the cost of public goods. The fragmentation of countries also resulted in some wasteful investment as small countries developed their own ports and transit corridors to neighboring landlocked countries. The duplication could be very costly, as in the case of Guinea (box 1.1). These difficulties are exacerbated by mutual suspicions, which prevent sensible economic collaboration. In the transport sector, the previously integrated railway administrations of Mali and Senegal and of Burkina Faso and Côte d'Ivoire were separated, to the detriment of all four coun- tries. Cross-border transport is particularly affected by such separations. Some of these problems have been overcome. The joint concessioning of separated railways has enabled them to be operated once again as unitary systems. And the creation of regional economic communities has enabled the development of some sound regional policies--especially with respect to the liberalization of international air transport. But there is still much to do to overcome these difficulties. State enterprises were excessive and inefficient. The commitment of many of the new leaders to Marxism has already been noted. Unfortunately, almost without exception these leaders had unrealistic expectations of what could be achieved with state ownership or control. Price controls and other populist impositions starved governments of cash and even- tually drove many of the enterprises into decline or bankruptcy. Many of the experiments with state ownership failed because of politicized management, with senior management appointments made on the basis The Legacy of History 5 Box 1.1 The Economic Costs of Political Fragmentation: The Case of Guinea The recent discovery of large iron ore deposits in Guinea by Rio Tinto Zinc raises important and difficult issues common in the postcolonial context. The exploita- tion of the deposits evidently requires investment in a mine, but the pertinent is- sue is the investment needed in transport infrastructure. A railway already links the deposit to a deepwater port, Buchanan, a legacy from the age of empires. But Buchanan is in Liberia, and the government of Guinea does not want to find its work captive to administrative holdups by the Liberian government. It has there- fore insisted that the transport of the iron ore be done entirely within Guinea, which requires the construction of a new dedicated railway and deepwater port. This decision has more than doubled the total investment needed for the project, adding around $4 billion.1 Evidently, these additional costs will be passed on to the people of Guinea. The government has agreed with Rio Tinto Zinc to absorb them through a reduced flow of royalty payments. The decision is also costly for the peo- ple of Liberia: the port of Buchannan is losing what may prove to be a key oppor- tunity for a scale economy. Source: Collier and Venables 2008. of political support or tribal and family membership, rather than on the basis of technical and managerial competence. The transport sector was not unique in this respect, as even the most ambitious of the industrial- ization programs failed. Corruption was rife. The wish to control Africa's rich natural resources was the major driver of colonialism. Independence redirected the gains not to the national populace but to its political leaders. In practice, one-party rule in Africa resulted in repression of minorities and extreme exploitation of national wealth by rulers. Control of the extraction and export of raw materials proved a major source of wealth for those who governed post- colonial Africa, as well as the root cause of several regional wars. A large proportion of the proceeds from developing mineral reserves was conspic- uously consumed by the rulers and their close associates at home or went into their bank accounts abroad. Meanwhile, domestic economies bene- fited little. A preparatory document for the African Union draft conven- tion on corruption in September 2002 estimated that corruption cost 6 Africa's Transport Infrastructure Africa $148 billion per year, more than 25 percent of the continent's gross domestic product (GDP) (African Union 2002). The corruption that will later be identified as a major source of inefficiency in the transport sys- tem is thus an expression of a general political malaise rather than any- thing specific to the transport sector. Civil wars were common. Military overthrow of corrupt civilian admin- istrations rarely eliminated the corruption (World Bank 1989). Spurred by events in Eastern Europe, many countries returned to multiparty pol- itics in the early 1990s, but doing so rarely eliminated corruption, and in several cases it unleashed historic ethnic hatreds in genocidal frenzies. By 2000, there were more than 10 major conflicts going on in Africa, and more than one-fifth of the total population lived in war-torn countries. These "fragile states" are among the poorest of nations and often have the worst transport facilities, as transport links, particularly railways and bridges, are prime targets in civil wars. The sector was poorly prepared for urbanization. This poor preparation had a number of root causes. The emphasis on rail rather than road devel- opment meant that urban road systems were often inadequate in density, badly constructed, and poorly maintained. The poor management of the state or municipal bus companies, together with attempts to maintain uneconomically low fares without any compensating subsidies, destroyed many of the conventional bus companies. The lack of adequate urban reg- ulatory institutions meant that the informal sector services that emerged were effectively subject only to self-regulation by operators' associations, which acted primarily in operators' rather than passengers' interests. The Outcome: High Costs, Poor Service, and Reduced Trade Inland transport costs in Africa are much higher than those in any other region of the world. The United Nations Conference on Trade and Development estimated that international transport costs faced by African countries, at 12.6 percent of the delivered value of exports, were more than twice as high as the world average of 6.1 percent (UNCTAD 2003). The United Nations Economic Commission for Africa put the average at 14 percent of the value of exports--and higher still for the landlocked countries such as Malawi (56 percent), Chad (52 percent), and Rwanda (48 percent)--compared with 8.6 percent for all developing countries (UNECA 2004). Moreover, freight moves slowly and uncertainly. Naudé and Matthee (2007) estimate that the reduction in trade resulting from this transport performance could be well in excess of 20 percent. The Legacy of History 7 Why is performance so bad? Mainly because political and economic conditions in Africa have prevented the development of the type of mod- ern logistical systems that have fostered trade and economic growth in the industrial world. The following aspects of logistics performance are encapsulated in the World Bank's Logistics Performance Index (LPI): · Efficient clearance of customs and other border-control agencies · High-quality information technology systems · Easy and affordable arrangement of shipments · Competence among transport operators, customs brokers, and so on · Ability to track and trace shipments · Adequate infrastructure (local transportation, terminal handling, warehousing) · On-time arrival Individual country performance is illustrated in map 1.1, which shows that, with the exception of South Africa, African countries south of the Sahara perform very poorly on aggregate. Looking at the separate components of the index makes clear that all countries except South Africa performed poorly not only on infrastructure quality but also on all the main aspects of logistics competence. Map 1.1 World Logistics Performance Index 1.00 LPI 2.48 2.48 LPI 2.75 2.75 LPI 3.23 3.23 LPI 5.00 no data Source: World Bank 2010. Note: LPI = logistics performance index; 1 is the minimum and 5 the maximum score. 8 Africa's Transport Infrastructure Several of these components relate to the efficiency of transport infrastructure in meeting the demands of tightly organized trading chains. The strength of those trading chains can be no greater than that of their weakest links, usually the interchanges. The weaknesses are partly physical--for example, in cases where there is a missing connec- tion between the modes or infrastructure needed for transshipment. They are partly institutional--as in cases where responsibility for the interchanges does not fall clearly with one agency or another. And they are partly operational--as when government's interest in collecting taxes and duties, or staff's interest in collecting bribes, slows down movement and drives up costs. The port-rail connection is the first major weakness. The compara- tive advantage of using rail over roads for long-distance transport of time-insensitive commodities means that railways depend heavily on international trade. Good rail-port connections are essential to com- plete the journey of goods overseas, but such connections are often inhibited by conflicts between rail and port authorities over control of rail movements in port areas; except in South Africa, inland transport and supporting facility investments are poorly aligned with port devel- opment. The stripping and stuffing of containers in port areas also cre- ates inland transport congestion in many ports. It is no accident that some of the most successful lines in Africa perform well in national cor- ridors where specialized rail and port facilities are vertically integrated (for example, the South African Transnet Freight Rail coal and ore lines and the Gabonese manganese ore line). Links among complementary rail systems are also essential. Some rail- way organizations have already created such links. The binational rail- ways in Burkina Faso­Côte d'Ivoire and Mali­Senegal offer the prospect of freer movement, as does the involvement of the same contractor in contiguous railways (the NLPI role in the route from South Africa through Zimbabwe to Zambia) or the same concessionaire (Central East African Railways Company in Malawi and Mozambique). But these arrangements also create local monopolies that can use predatory prac- tices to increase profits, as in the case of the Zambian treatment of Congolese copper exports (see chapter 3 of this book). In East Africa, joint concessioning of railways is part of a World Bank­funded corridor, including the reform of border-crossing arrangements. Some countries are now trying to develop coordinated corridor systems, as in the Ghana Gateway and Maputo corridors. The Legacy of History 9 Whatever the mode of transport, however, the most serious impedi- ments are administrative. For road transport, the regulation and market structures of the road freight industry, rather than the quality of road infrastructure, are the binding constraints on international corridors (Teravaninthorn and Raballand 2008). Third-party logistics, which have played such a large role in increasing production and distribution efficiency in industrialized countries, are still poorly developed in Africa. Customs and transshipment improvements are also central to corridor improvement. Some landlocked countries already have bonded warehouses at ports in West Africa. Concessionaires are also speeding transit, such as through the Sitarail intermodal terminal proposal in Ouagadougou, the Zambia Rail company customs bond at Victoria Falls, and the planned Madarail bonded container terminal near Antananarivo. There is scope for a regional program on trade facilitation similar to the successful effort in southeastern Europe, which was catalyzed by the prospect of entry into the European Union. Transport in Africa is also very unsafe. Vehicles and infrastructure are poorly maintained. Failures of governance accentuate the problem, as policing is corrupt and laws are not enforced. Over the past two decades, life in general has been precarious and violence prevalent. In such circum- stances, transport safety is not an obvious priority. It is therefore not sur- prising that all modes of transport, in particular road and air, have extremely poor safety records. While the nature and causes of incidents differ between these two sectors--as will be discussed in more detail-- neither mode has a developed safety culture. As with so many aspects of transport in Africa, the problem is deeply embedded in the continent's recent troubled history. General social sta- bilization should help, but deeply ingrained attitudes are difficult to change. Unless such attitudes are overcome, no amount of infrastruc- ture development is likely to bring about much improvement. The necessity of transport infrastructure for economic development is taken as axiomatic. But transport infrastructure in Africa is judged insuf- ficient for achieving this end in two important senses. First, and most obviously, the region is found to be quantitatively underendowed com- pared with other regions of the world. Its road system is less dense, its rail system built to lower standards, its ports ill-equipped for the develop- ment of containerization, and its air transport system lacking in adequate air traffic control and navigation services. Second, the region's physical infrastructure is not accompanied by good transport service; that is, the infrastructure is not well maintained, managed, or operated. Hence, it is 10 Africa's Transport Infrastructure not only in physical terms but also in governance that transport infra- structure in Africa is insufficient. Country Diversity and Uneven Economic Performance Despite similarities in their postcolonial political history and problems, Africa's many countries face diverse economic conditions. Understanding that structural differences in economies and institutions affect countries' growth and financing challenges as well as their economic decisions (Collier and O'Connell 2006), this book categorizes the nations studied into four types to organize much of the discussion (map 1.2). Map 1.2 A Country Typography of Sub-Saharan Africa CAPE MAURITANIA VERDE MALI NIGER CHAD ERITREA SENEGAL THE GAMBIA SUDAN BURKINA GUINEA-BISSAU GUINEA FASO BENIN GHANA NIGERIA ETHIOPIA TOGO SIERRA LEONE CÔTE D'IVOIRE CENTRAL AFRICAN REPUBLIC SOMALIA LIBERIA CAMEROON EQUATORIAL GUINEA UGANDA SÃO TOMÉ AND PRÍNCIPE CONGO DEMOCRATIC KENYA GABON REPUBLIC RWANDA OF CONGO BURUNDI TANZANIA SEYCHELLES COMOROS ANGOLA MALAWI Resource ­ Rich ZAMBIA Low Income Country ­ Not Fragile MOZAMBIQUE MADAGASCAR ZIMBABWE Low Income Country ­ Fragile NAMIBIA ANA BOTSWANA MAURITIUS Middle Income Country SWAZILAND SOUTH LESOTHO AFRICA Source: Briceño-Garmendia, Smits, and Foster 2009. The Legacy of History 11 The categories shown in map 1.2 are defined as follows: · Middle-income countries have GDP per capita in excess of $745 but less than $9,206. Examples include Cape Verde, Lesotho, and South Africa (World Bank 2010). · Resource-rich countries are low-income countries whose behaviors are strongly affected by their endowment of natural resources (Collier and O'Connell 2006). These countries typically depend on minerals, petroleum, or both. A country is classified as resource rich if primary commodity rents exceed 10 percent of the GDP. South Africa is not classified as resource rich, using this criterion. · Low-income, fragile states face particularly severe development chal- lenges, such as weak governance, limited administrative capacity, vio- lence, or the legacy of recent conflict. Countries that score less than 3.2 on the World Bank's Country Policy and Institutional Performance Assessment (WBI 2004) belong to this group. Some 14 African coun- tries are in this category. · Low-income, nonfragile states are those that have GDP per capita below $745 and are neither resource rich nor fragile. Table 1.1 shows how all the countries in the study are categorized and also notes whether they are coastal or landlocked. The most significant feature of Africa is that it has a much larger pro- portion of low-income countries than the rest of the developing world. An important element of this poverty is that nearly one-third of African countries are classified as low-income, fragile states that have recently suf- fered from major political and economic trauma. But even its resource- rich countries have an average GDP per capita that is only 70 percent of that of the lower-middle-income developing countries in the rest of the world (table 1.2). African countries share some common economic features. Thirty- seven percent of their populations lives in cities, with little variation among the four country types. Agriculture accounts for about a third of the GDP, on average, again with relatively little variation among the country types. But other features are not shared so equally: for example, the share of land available for agriculture varies from a low of 29 percent for the low-income, fragile states to 63 percent for the middle-income countries. The trade share of GDP ranges even more widely, from 120 percent for low-income countries to 39 percent for the low-income, fragile states (World Bank 2009).2 Together with 12 Table 1.1 Typology of Countries Alternative classification for low-income countries Low-income, Low-income, Low-income, Low-income, Resource-rich nonfragile fragile Middle-income coastal landlocked Angola Benin Burundi Botswana Benin Burkina Faso Cameroon Burkina Faso Central African Cape Verde Comoros Burundi Chad Ethiopia Republic Lesotho Côte d'Ivoire Central African Congo, Rep. Ghana Comoros Mauritius Eritrea Republic Equatorial Guinea Kenya Congo, Dem. Rep. Namibia Gambia, The Congo, Dem. Rep. Gabon Madagascar Côte d'Ivoire Seychelles Ghana Ethiopia Nigeria Malawi Eritrea South Africa Guinea Malawi Sudan Mali Gambia, The Swaziland Guinea-Bissau Mali Mozambique Guinea Kenya Niger Niger Guinea-Bissau Liberia Rwanda Rwanda Liberia Madagascar Uganda Senegal São Tomé and Mozambique Zambia Tanzania Príncipe São Tomé and Zimbabwe Uganda Sierra Leone Príncipe Zambia Togo Senegal Zimbabwe Sierra Leone Tanzania Togo Source: Classification as proposed in IMF (2007), except for Sudan, which was added for completeness. Table 1.2 Basic Characteristics of African Countries vs. World's Other Developing Countries Africa Rest of the world Low-income, Low-income, Middle- Lower-middle- Upper-middle- Total Resource-rich nonfragile fragile income Low-income income income Number of countries 48 9 15 15 9 13 44 39 GDP per capita, PPP (constant 2005 US$) 843 930 342 269 4,850 421 1,337 5,329 Population (millions) 763 239 324 143 56 423 3,300 874 Land area (millions of km2) 23.6 7.6 8.5 4.9 2.7 3.7 25.5 43.3 Total GDP (constant 2005 US$ billions) 643 222 111 39 271 178 4,413 4,655 Urban population as share of total (%) 34.9 45.5 23.8 34.5 55.8 28.9 38.7 74.7 Trade as share of GDP (%) 69.8 82.7 59.7 84.7 61.3 75.9 68.1 58.2 Agricultural land as share of total land (%) 44.0 48.3 42.5 29.3 63.3 38.8 49.8 28.1 Source: Carruthers, Krishnamani, and Murray 2009. Note: km2 = square kilometer; PPP = purchasing power parity. 13 14 Africa's Transport Infrastructure geographic and climatic features, these similarities and differences--which are even greater when individual countries are considered--contribute to the varying needs for and costs of transport infrastructure. A New-Millennium Renaissance Fortunately, the portents are not all ominous. The advent of democracy in South Africa in 1994 acted as a catalyst for some change. The launch of the New Partnership for Africa's Development by a group of 15 African states in 2001 and the replacement of the Organization of African Unity by the African Union in 2002 heralded a commitment not only to more democratic political processes but also to multilateral action as a means of achieving and maintaining them. The Clinton initiative in 1998 and the Commission for Africa report in 2005 reflected the sup- portive Western attitude. Hopefully, these mark the beginnings of a political renaissance. Partly as a consequence of these developments there has been at least a minor renaissance in Africa's economy. From 2004 to 2008, it expanded by more than 5 percent every year--the first time in more than 45 years that such a growth rate had been sustained over a long period.3 In 2008, the overall growth rate was 5.4 percent despite the global economic downturn (World Bank 2009). These figures suggest the emergence of an economic renaissance. The outlook for the immediate future is also promising in spite of the poor world economic climate. Admittedly, weaker external demand and lower commodity prices will take a toll. In particular, declines in demand in key external markets are likely to lead to a negative trend in the contribution of trade to GDP growth, with an impact on inter- national transport demand, particularly for shipping services (discussed in chapter 5). Official development assistance flows may also slow, which is of particular significance when considering how fast backlogs in capital investment can be overcome, as discussed in chapter 8. Nevertheless, growth is forecast to slow only to 3.5 percent overall. This is partly because the African economies are not well integrated into the international financial system. Hence the direct effects of the global finan- cial and economic crisis were considered likely to be limited in the African economies, according to a World Bank Global Economic Prospects review in mid-2009 (World Bank 2009). In summary, Africa has inherited from its history a distorted and rela- tively poor transport infrastructure, which it has neither managed nor The Legacy of History 15 maintained well. It has been heavily dependent on official development assistance for much of its transport spending but has nonetheless achieved substantial economic growth and has prospects for more. Against this background, the book now moves on to discuss the major modes of transport. Notes 1. Throughout the book, monetary values are given in U.S. dollars unless other- wise specified. 2. Trade share of GDP measures the importance of trade to an economy. Merchandise trade as a share of GDP is the sum of merchandise exports and imports divided by the value of GDP, all in current U.S. dollars. According to the World Bank (2009), the highest ratio in 2005 was that of Singapore, with a value of 368, while Hong Kong, China, had a value of 333. Equatorial Guinea and Liberia ranked fourth and fifth with values of 285 and 253. 3. The main exception was South Africa, which grew by only 3.4 percent and appeared to be facing weaker demand for its exports (World Bank 2009). References African Union. 2002. "Draft African Convention on Preventing and Combating Corruption." Document prepared at the Ministerial Conference of the African Union, Addis Ababa, September 18­19. Briceño-Garmendia, C., K. Smits, and V. Foster. 2009. "Financing Public Infrastructure in Sub-Saharan Africa: Patterns, Issues and Options." Africa Infrastructure Country Diagnostic Background Paper 15, World Bank, Washington, DC. Carruthers, R., R. R. Krishnamani, and S. Murray. 2009. "Improving Connectivity: Investing in Transport Infrastructure in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 7, World Bank, Washington, DC. Collier, P., and S. A. O'Connell. 2006. "Opportunities and Choices." In The Political Economy of Economic Growth in Africa 1960­2000, ed. B. J. Ndulu, S. A. O'Connell, J.-P. Azam, R. H. Bates, A. K. Fosu, J. W. Gunning, and D. Njinkeu. Cambridge: Cambridge University Press. Collier, P., and A. Venables. 2008. "Trade and Economic Performance: Does Africa's Fragmentation Matter?" Paper presented at the Annual Bank Conference in Development Economics, World Bank, Washington, DC, May 28. IMF (International Monetary Fund). 2007. World Economic Outlook. Washington, DC: IMF. 16 Africa's Transport Infrastructure Meredith, M. 2005. The State of Africa. London: Free Press. Naudé, W., and M. Matthee. 2007. "The Significance of Transport Costs in Africa." UNU Policy Brief 5, United Nations University­World Institute for Development Economics Research, Helsinki. Teravaninthorn, S., and G. Raballand. 2008. Transport Prices and Costs in Africa: A Review of the Main International Corridors. Washington, DC: World Bank. UNCTAD (United Nations Conference on Trade and Development). 2003. Efficient Transport and Trade Facilitation to Improve Participation by Developing Countries in International Trade. Geneva: UNCTAD. UNECA (United Nations Economic Commission for Africa). 2004. Assessing Regional Integration in Africa. New York: UNECA. WBI (World Bank Institute). 2004. Building State Capacity in Africa: New Approaches, Emerging Lessons. Washington, DC: WBI. World Bank. 1989. From Crisis to Self-Sustainable Growth. Washington, DC: World Bank. ------. 2009. Global Economic Prospects. Washington, DC: World Bank. ------. 2010 Logistics Performance Index (software). Washington, DC: World Bank. http://go.worldbank.org/88X6PU5GV0. CHAPTER 2 Roads: The Burden of Maintenance Roads dominate the transport sector in most African countries, carrying 80 to 90 percent of passenger and freight traffic. Moreover, they are the only means of access to most rural communities. This dominance is achieved despite the fact that the density of the region's network is lower, both per person and per square kilometer of land area, than that of other world regions. The condition of the road system is also poor by interna- tional standards. Nevertheless, the fiscal burden of the road network per capita is rela- tively high--a consequence of the combination of low population den- sity and low gross domestic product (GDP) per capita. In these difficult circumstances, the provision of secure funding for road maintenance and efficient implementation of that maintenance are critical to the effec- tiveness of the sector. While reforms in both of these areas over the past 15 years have improved performance, there is much left to do. In addition, road use needs to be efficient. Unfortunately, much remains to be done in this area also. The road freight industry is heavily cartelized and controlled, and yields high profits despite high costs. Road passenger transport--particularly in urban areas--has suffered from counterproductive fare regulation, with the result that most service is now provided by an informal sector that is largely self-regulated. 17 18 Africa's Transport Infrastructure This chapter analyzes more fully the nature and performance of the African road networks and their main commercial user, the road freight sector.1 It is based on three data sources: (i) a comprehensive road net- work survey undertaken specifically for the Africa Infrastructure Country Diagnostic (AICD), (ii) an institutional database prepared and main- tained as part of the Sub-Saharan Africa Transport Policy Program (SSATP), and (iii) a fiscal cost study undertaken as part of the AICD (appendix 2a). The basic country data--including land area, population, GDP, vehicle fleet, and transport fuel consumed--are shown in appendix 2b.Two kinds of country typologies are used to facilitate the presentation of the results. The first relates to factors completely exogenous to the road sector but that could nonetheless be expected to influence it significantly. These fac- tors include macroeconomic circumstances (countries are classified as middle-income, low-income, or resource-rich; or as low-income and aid- dependent),2 geography (coastal, landlocked, or island), and terrain (flat and arid versus rolling and humid). The second set of factors relates to policy variables, which are completely endogenous to the road sector. These factors include institutions (namely whether the country has a road fund, a road agency, or both) and funding mechanisms (for example, the existence of a fuel levy and the level at which it is set). The Road Network The size of the classified road network, including the main roads and sec- ondary network, is estimated to be 1,052,000 kilometers (km). Together with an unclassified network of 492,000 km and an urban road network of about 193,000 km, this makes an estimated total network of 1,735,000 km (appendix 2c). Strategic Roads: Serving International Transit Corridors Relatively few international road transport corridors play a crucial role in maintaining the economies of the landlocked countries of Africa. Of these, the main international trade corridors that connect the landlocked countries of each subregion to their respective ports are widely consid- ered the most important. Some $200 billion worth of imports and exports per year move along these key corridors, which have a combined length of little more than 10,000 km (table 2.1). For Central Africa, regional transport is dominated by two road and rail corridors, which link the port of Douala in Cameroon with Chad Roads: The Burden of Maintenance 19 Table 2.1 Overview of Africa's Key Transport Corridors for International Trade Length Roads in good Trade density Implicit speed Freight tariff Corridor (km) condition (%) (US$ million/km) (km/hour) (US$/tonne-km) Western 2,050 72 8.2 6.0 0.08 Central 3,280 49 4.2 6.1 0.13 Eastern 2,845 82 5.7 8.1 0.07 Southern 5,000 100 27.9 11.6 0.05 Source: Adapted from Teravaninthorn and Raballand (2008). Note: Implicit speed includes time spent stationary at ports, border crossings, and other stops. (serving cotton and oil exports) and the Central African Republic (serv- ing logging exports). For West Africa, there are several potential gateways (in Benin, Côte d'Ivoire, Ghana, Guinea, Senegal, and Togo) serving the landlocked coun- tries of Burkina Faso, Mali, and Niger. But the closing of the international routes from Abidjan as a consequence of the crisis in Côte d'Ivoire has meant that most of the traffic now goes through ports in Benin, Ghana, Togo, and with Burkina Faso also becoming a transit country for Mali. Some 50 percent of the import traffic to Burkina Faso is now routed through Lomé, Togo, and 36 percent through Tema, Ghana. In East Africa, 80 percent of trade flows originate or terminate outside the region, despite the creation of the East African Community (EAC). Mombasa is the main port for the region, handling more than 13 million tonnes of freight per year and serving not only Kenya and Uganda but also Burundi, the Democratic Republic of Congo, and Rwanda through the northern corridor. The central corridor from Dar es Salaam also serves the Democratic Republic of Congo as well as being an alternative for Zambia. In southern Africa, there are four significant trade routes. The main route, the north-south corridor from Durban, serves as an intraregional trade route linking Zambia, southeastern Democratic Republic of Congo, and western Malawi with Botswana, Zimbabwe, and South Africa. The alternate routes through Beira, Walvis Bay, and Dar es Salaam, although closer to parts of the region, suffer relative to the north-south corridor from Durban, because of the latter's superior road infrastructure, better port equipment, and lower maritime rates. The idea of creating a comprehensive continental road system in Africa--the Trans-African Highway network--was formulated in 1970 as part of a political vision for pan-African cooperation. As envisioned, the system would consist of nine main corridors with a total length of 20 Africa's Transport Infrastructure 59,100 km. In any event, national governments have not committed the financing needed to make this network a reality. While many of the roads already exist as elements of national highway networks, almost half of the 50,000 km that could be used is in poor condition. About 70 percent is currently paved, but 25 percent has either an earth surface or is not devel- oped at all. Most of the missing links are concentrated in Central Africa. When the status of the concept was reviewed in 2003, it was found that of nine proposed links in the network, only one, Cairo­Dakar, was near complete (African Development Bank 2003). It was estimated that the costs of completing the whole network would be over $4 billion. At that price the network's future looks dubious. National Classified Roads: Too Sparse The spatial density of roads in Africa is low by international standards. The country-weighted average is 109 km of classified roads and 149 km of all roads per 1,000 square kilometers (km/1,000 km2) of land area, with median values of 57 and 82, respectively. With the exception of Mauritius, which has 993 km/1,000 km2, the classified road densities range between 10 (Mauritania) and 296 (The Gambia). For density per capita, the average total classified network density is 2.5 km per 1,000 people, and the median value is 1.5. But there is huge variation, with total network density as low as 0.5 km per 1,000 people in Burundi and Rwanda and as high as 21.0 in sparsely populated Namibia (appendix 2d). Overall, about one-quarter of the networks are designated as primary, one-quarter secondary, and one-half tertiary, with unclassified networks about equal to the tertiary. At one extreme, Lesotho, Namibia, and South Africa have around 50 km of primary roads per million people, while at the other extreme, Niger and Uganda have more than 1,000 km of primary roads per million people. The variation in secondary road den- sities is lower, with most countries having secondary network densities of between 10 km and 100 km per million (figure 2.1). In terms of road space per vehicle, the number of kilometers of classi- fied road per 1,000 vehicles ranges from 950 in the Central African Republic to only 11 in Nigeria, with a country-weighted average value of 152 and a median value of 82 km per 1,000 vehicles. The proportion of road that is paved also varies greatly (see appendix 2e). While on average 64 percent of primary roads and 17 percent of all clas- sified roads are paved, the richer countries such as South Africa and Botswana have a higher proportion paved. But three countries (the Central African Republic, Chad, and the Democratic Republic of Congo) 21 Et kilometers per 1,000 people 0 1 2 3 4 5 6 hi Rw opi Figure 2.1 an a Ni da g Bu eria ru Se nd ne i Su gal Co da ng Ni n o, ge D T r M em ogo oz . am Re bi p. q M G ue ad ha ag na as c Be ar Za nin m bi primary network Ca Ken a Cô m ya te ero d' on Iv o Bu Gu ire Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. rk in Range of Primary and Secondary Road Densities in ea a Ta Fas nz o a Si M nia er al ra aw Le i on secondary network Zi Erit e m re Ga b a m ab bi we a, Ug The an da M Ch au ad ri An tius go Lib la er ia M Co Les al ng ot i percentage primary Ce So o, ho nt ut Re ra l A M h A p. fri a fri ca ur ca n ita Re ni Sw pub a az lic ila G nd Bo abo ts n w Na ana m ib ia 0 10 20 30 40 50 60 70 80 90 100 secondary network as percentage of primary network 22 Africa's Transport Infrastructure have less than 20 percent of their primary network paved, and more than one-quarter of the countries have 10 percent or less of their total road network paved. Rural Transport Infrastructure: Critical to Agriculture Rural transport infrastructure consists of more than designated and mapped roads. In rural areas, people and vehicles move across myriad unclassified and unrecorded paths and tracks. The size of the rural net- work is thus difficult to determine. But it is estimated that Africa has about 1 million km of designated rural roads (either tertiary or unclassi- fied), whose replacement value is estimated at $48 billion, together with a network of undesignated rural roads, tracks, paths, and footbridges, which may be one and a half to two times as extensive as the local gov- ernment road networks. In most countries, the majority of rural network kilometers are cap- tured by the official tertiary network. But in a number of cases--includ- ing Benin, Ethiopia, and Rwanda--less than one-third of the rural network is classified. Figure 2.2 shows the huge variation in the density of this rural road network as well as the relative weight of classified terti- ary roads. The density of rural roads (tertiary and unclassified) ranges from 0.1 km per 1,000 people in the Democratic Republic of Congo to 21.6 in Namibia, with a mean of 2.6 km and a median of 1.2 km. Burkina Faso, Namibia, and South Africa stand out as having extensive rural net- works relative to their populations. A low density of rural roads limits access to agricultural production, which accounts for one-third of the region's GDP and 40 percent of its export revenues. The Region's Roads in an International Context Africa has a much lower spatial density of roads than any other region of the world (figure 2.3). It has only 204 km of roads per 1,000 km2 of land area, with only one-quarter paved, while the world average is 944 km/1,000 km2, with over half paved. The spatial density of Sub- Saharan Africa's roads is less than 30 percent that of South Asia, where half of the roads are paved, and only 6 percent that of North America, where two-thirds are paved. To some extent, this low spatial density reflects the low population densities of Africa. Sub-Saharan Africa has a total road network of 3.40 km per 1,000 people, compared with a world average of 7.07 km (figure 2.4). The road density with respect to population in Sub-Saharan Africa is actually slightly higher than that of South Asia, which has 3.19 km of 23 Co ng o, kilometers per 1,000 people De m Figure 2.2 0 5 10 15 20 25 .R M ep. a Et law hi i o S pia M ud au an Ta ritiu nz s an N ia Bu ige ru r Bu E nd rk rit i Cô in re te a F a d' aso Iv oi Si r er To e ra g Le o o Ke ne tertiary network Ni nya g M Rw eri oz a a am nd Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. bi a q M Ug ue ad an Range of Tertiary and Unclassified Road Densities ag da a Se sca ne r Le ga Sw sot l az ho ila unclassified network Be nd n Gh in an C a G h Ca uinad m e M er a a o Ga uri on m tan bi ia a, T Zi Mhe Ce m a nt ba li ra b lA Za we m percentage tertiary fri ca A bia n ng Re o pu la Co L blic n ib So go, eria ut Re h p Bo Afr . ts ica w Ga ana Na bon m ib ia 0 10 20 30 40 50 60 70 80 90 100 tertiary as percentage of total rural network 24 Figure 2.4 Figure 2.3 km/1,000 people km/1,000 km2 0 5 10 15 20 25 30 Su W 0 500 1,000 1,500 2,000 2,500 3,000 3,500 b- Su Source: World Bank 2009. Source: World Bank 2009. Sa or ld b- W ha Sa or ra ha ld M n ra id Af ric M n Af dl e a id ric Ea So dl e a st ut Ea So La an h st ut h Africa's Transport Infrastructure tin d As La an Am No ia tin d As rth Am No ia er ic aa Af er rth ric ica Af nd a an ric Ea Ca d a st rib Ea Ca As be st rib Eu ia an As be ro an Eu ia an pe d ro an an Pa pe d d cif an Pa Ce ic d cif nt Ce i c ra nt lA ra Total Road Networks per Capita in World Regions No sia lA Spatial Density of Road Networks in World Regions rth No sia Am rth er Am ic a er ica Roads: The Burden of Maintenance 25 roads per 1,000 people, and only slightly lower than that of the Middle East and North Africa, which has 3.88 km per 1,000 people. But the paved road length in Sub-Saharan Africa, 0.79 km per 1,000 people, still remains less than half of that of South Asia and only about one-fifth of the world average. Moreover, given low GDP, the fiscal burden of maintaining this limited road network is significantly higher than elsewhere (figure 2.5). Sub- Saharan Africa has a total road network of 6.55 km per $1 million, com- pared with South Asia's 5.32 and a world average of 3.47. The North American equivalent value, 0.79 km per $1 million, is just over a tenth that of Sub-Saharan Africa. With respect to paved roads, Africa has a network of 1.12 km per mil- lion dollars of GDP, which is only slightly higher than the world average of 0.98, and less than South Asia's average of 2.67. Table 2.2 compares the paved road networks of the AICD countries with those of other lower-income and lower-middle-income countries of the world. It shows that lower-income countries in Africa have lower levels of paved roads per capita, per square kilometer, and per GDP per capita than other low- income countries in the world. While African low-income countries have lower average population densities than low-income countries in the rest Figure 2.5 Total Road Network as Share of GDP in World Regions 7 6 5 km/million US$ 4 3 2 1 0 ia a an ic sia a ld a ric ic ric cif As or be lA er Af Af Pa W h Am rib ra ut rth an d nt Ca So an r rth No Ce ha d sia No Sa an d d an an tA b- ica Su s e st Ea op er Ea Am r Eu e dl tin id M La Source: World Bank 2009. 26 Africa's Transport Infrastructure Table 2.2 Cross-Regional Comparison of Paved Road Infrastructure in Low-Income Countries Paved roads Units Africa Rest of world Density by area km/1,000 km2 10.7 37.3 Density by population km/1,000 people 269.1 700.7 Density by GDP per capita km/US$ billion 663.1 1,210.0 Source: Carruthers, Krishnamani, and Murray 2009. of the world (70 per km2 compared with 125), the relative disparity in the proportion of paved roads is substantially greater than this (10.7 km per 1,000 km2 compared to 37.3 km). Road Traffic Volumes: Manageable . . . for Now Traffic volumes in Africa are relatively low by international standards (appendix 2f). The annual average daily traffic on roads in the primary net- work ranges from only 50 vehicles in the Democratic Republic of Congo and the Republic of Congo to over 7,000 in Mauritius and slightly less in South Africa. The country-weighted average is 1,198 and the median 829. Of the larger countries, only Nigeria and South Africa have heavy average volumes on the main road network (figure 2.6). Such low volumes effec- tively preclude the possibility of financing roads from tolls in most coun- tries. Volumes on the secondary networks range from 746 vehicles per day in Nigeria to less than 30 in seven countries, with a country-weighted mean of 185 and a median value of 126. Traffic is heavily concentrated on the main road network (see appen- dix 2g). In most countries, at least 90 percent of reported traffic on the classified network is carried on the main networks, which typically com- prise centrally administered primary networks plus secondary networks. But in a handful of countries (Malawi, Nigeria, South Africa, and Uganda), only the primary network is centrally administered. Rural networks typically carry very low levels of traffic, amounting to no more than 10 percent of overall traffic on the classified network (figure 2.7). In a handful of countries, the rural network plays a more prominent role, capturing more than 20 percent of traffic--namely, in Ethiopia, Malawi, and Nigeria. But with the exception of Nigeria, the absolute vol- umes of traffic on the rural network are very low, averaging around 30 vehicles per day. Fourteen countries have an average daily traffic rate of fewer than 10 vehicles on their tertiary networks (appendix 2h). The 27 percentage of total traffic Figure 2.6 Et 0 10 20 30 40 50 60 70 80 90 100 hi o M pia al Co Ni awi ng ge o, ria M U Rep oz ga . Ce am nd nt bi a ra qu lA Ni e fri ca Za ger n m Co Re bi ng pu a o, Se bli De ne c m ga .R l Na ep m . i Ke bia ny Ch a a Ca Gh d m an er a Volume of Traffic Carried on Main Network Bu L oon rk ibe in ri a a Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. traffic share Bo Fa ts so w M Les ana ad o ag tho as ca Cô M r te d Su ali d' an Iv So G oire ut ab Ga h A on m f Si bia rica er , ra Th Le e average annual daily traffic Gu one Ta in M nza ea au n r ia M itan au ia rit i Be us B n Zi uru in m n ba di b An we Rw gol an a Er da itr e Sw To a az go ila nd 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 average annual daily traffic (vehicles per day) 28 Sw percentage of total traffic Figure 2.7 az 0 10 20 30 40 50 60 70 80 90 100 ila n To d Er go Rw itre a a A nd Zi ng a m ol ba a Bu bw ru e n M n Be di a M uri in au tiu r s Ta itan nz ia a Si G nia er ui ra ne Ga L a m eo So bia ne ut , T h he A Cô G fric te ab a Volume of Traffic Carried on Rural Network d' on Iv o Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. Su ire traffic share da M ad M n ag al a i Le sca Bo sot r Bu ts ho rk wa in n aF a as Ca Lib o m eri er a o Gh on an Ch a Co a Ce ng N Ken d o, a ya average annual daily traffic nt De mi ra lA m bia fri . ca Se Rep n ne . Re g pu al Za blic m M b oz N ia am ig bi er q Co Ug ue ng an o, da R Ni ep ge . M ria a Et law hi i op ia 0 100 200 300 400 500 average annual daily traffic (vehicles per day) Roads: The Burden of Maintenance 29 highest average for the tertiary network is in Mauritius, with 200 vehicles per day. Road Infrastructure Performance Ideally, the performance of the road infrastructure should be measured in terms of the speed, cost, and safety of traffic using it. However, as detailed measurement of these variables is not available, the measured condition of roads is used instead as a proxy. Condition of Main Roads: Poor, but Improving Road condition is the primary indicator of the performance of a road management system. In the AICD study, link-by-link data were collected on the quality of the sample countries' main road networks (managed by the central government or affiliated agency) and rural networks (managed by local governments) (see appendix 2i). A three-way quality classifica- tion was used: good, fair, and poor. "Poor" designates roads in need of reha- bilitation. The data on the rural networks are less reliable than those on the main road networks, as subnational field visits were not made. Figure 2.8 shows huge variation in the percentage of main roads in good condition but slightly less variation in the percentage of main roads in good or fair condition. On average, about 43 percent of the main net- works are in good condition, a further 31 percent are in fair condition, and the remaining 27 percent are in poor condition.3 The percentage in good condition ranges from 4 percent in the Republic of Congo to 90 percent in South Africa. But the percentage in good or fair condition covers a narrower range, from 27 percent in the Republic of Congo to 98 percent in South Africa. In five countries, more than 50 percent of the primary network is in poor condition (Democratic Republic of Congo, Republic of Congo, Guinea, Senegal, and Togo), while in six (mostly mid- dle-income) countries, less than 10 percent of the primary network is in poor condition (Burkina Faso, Mauritius, Namibia, South Africa, Swaziland, and Tanzania). Of particular interest are the trends in road quality over time. Unfortunately, time-series data on road conditions are extremely limited. An early detailed review of new, "second-generation" road funds showed improvements in outcomes for the five countries (Benin, Ethiopia, Ghana, Malawi, and Zambia) for which road condition data were avail- able (Gwilliam and Kumar 2003). More recent trends in road conditions have also been broadly positive. But there are only very limited data 30 Africa's Transport Infrastructure Figure 2.8 Distribution of Road Network Length across Condition Classes a. Main network South Africa Mauritius Burkina Faso Botswana Swaziland Central African Republic Kenya Namibia Burundi Ethiopia Malawi Ghana Zambia Mozambique Niger Mauritania Cameroon Gabon Mali Nigeria Sierra Leone Madagascar Tanzania Angola Chad Senegal Sudan Zimbabwe Lesotho Benin Gambia, The Rwanda Congo, Dem. Rep. Eritrea Togo Guinea Uganda Liberia Côte d'Ivoire Congo, Rep. 0 25 50 75 100 percentage of road network length good fair poor Roads: The Burden of Maintenance 31 b. Rural network Mauritius Burkina Faso Ghana Kenya Malawi Mauritania Swaziland Liberia Guinea Gabon Eritrea Central African Republic Burundi Togo Chad Mali Sierra Leone Namibia Cameroon Ethiopia Lesotho Niger Tanzania South Africa Botswana Congo, Dem. Rep. Côte d'Ivoire Nigeria Zimbabwe Congo, Rep. Benin Uganda Senegal Mozambique Sudan Gambia, The Angola Zambia Madagascar Rwanda 0 25 50 75 100 percentage of road network length good fair poor Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. 31 32 Africa's Transport Infrastructure available on road quality trends, collected by the SSATP for a sample of 16 countries between 2004 and 2007. While some of the figures for 2004 depend heavily on local engineers' judgments (as opposed to independ- ent technical assessments), and subsequent changes in classification reduce the reliability and comparability of the data, overall, many coun- tries appear to have made substantial progress in improving the quality of their main road networks. Half of the sample increased the percentage of main roads in good or fair condition by more than 10 points--and more than 30 points in a number of cases (table 2.3). Only in a handful of cases has there been a significant deterioration in quality (Côte d'Ivoire, Ghana, Guinea, and Lesotho). In the case of Côte d'Ivoire, this can be attributed to the general collapse of services associated with political unrest. Furthermore, the gap between the best and worst performers has been closing over time, from 90 percentage points in 2004 to around 50 per- centage points in 2007. The asset value of the classified road networks--in their current con- dition, as a percentage of the asset value of the same networks in entirely good condition--can be assessed with the Road Network Evaluation Tool (RONET) data analysis. The value of this indicator is strongly influenced Table 2.3 Trends in Road Condition, 2004­07 % of roads in good or fair condition Change to good or fair August 2004 September 2007 (percentage points) Madagascar 30 75 +45 Mali 44 80 +36 Burundi 5 40 +35 Tanzania 50 69 +19 Benin 75 93 +18 Chad 30 48 +18 Niger 57 72 +15 Kenya 57 67 +10 Cameroon 60 65 +5 Ethiopia 62 65 +3 Malawi 63 65 +2 Mozambique 70 72 +2 Guinea 66 62 ­4 Ghana 65 60 ­5 Côte d'Ivoire 60 50 ­10 Lesotho 96 80 ­16 Sources: SSATP 2004, 2007. Roads: The Burden of Maintenance 33 by the condition of the paved road networks, since these have a much higher replacement cost per kilometer than the unpaved networks. The range runs from a minimum of 62 percent for Togo to a maximum of 94 percent for Mauritius, with the majority of countries having scores in the 80 to 90 percent range (figure 2.9). These findings indicate that coun- tries are sensibly focusing their efforts on maintaining their high-value paved networks in good or fair condition. Quality of Rural Roads: Generally Poor As might be expected, the condition of the rural networks is substantially lower than that of the main road networks (figure 2.10). On a country- weighted average, about 33 percent of the tertiary road networks are in good condition, a further 23 percent are in fair condition, and the remain- ing 44 percent are in poor condition. The percentage in good condition ranges from zero in Rwanda to 77 percent in Mauritius. Only four coun- tries have more than 70 percent of their tertiary networks in good or fair condition (Burkina Faso, Ghana, Malawi, and Mauritius); the average is 56 percent and the median is 54 percent. There is a fairly strong correla- tion between the quality of the main road networks and the quality of the rural road networks in a given country (figure 2.11). This correlation sug- gests that there is a country effect, with competence in main network management carrying over into the rural networks. One way of assessing the performance of a rural network is to con- sider the level of accessibility it offers to rural inhabitants. This measure is encapsulated in the Rural Accessibility Index (RAI), which measures the proportion of the rural population within a two-kilometer walking distance of an all-season road (World Bank 2007a). Based on household survey evidence analyzed for 20 countries in Africa, the RAI has an aver- age value of less than 40 percent (World Bank 2007b), compared with an average of 94 percent for richer borrowing country members of the World Bank. Estimating the RAI is possible using a geographic information system (GIS) model of Africa's road network and the geographical distribution of population (figure 2.11). The average value of the estimated RAI was even lower than the surveyed value--only 22 percent for the 24 countries in the sample. Countries such as Ethiopia, Niger, Sudan, and Zambia show particularly low estimated RAI--under 20 percent. Even Namibia, with its extensive rural network, reaches an RAI of just over 20 percent. If one uses the same GIS model, it is possible to calculate how many kilometers of additional tertiary roads would need to be built to reach a 100 percent 34 Figure 2.9 percent Co C 0 20 40 60 80 100 ng on To o, go go De , R m ep. . Se Rep ne . Ug ga an l Gu da Le ine Ca sot a m ho er o Ga Za on m mb bi ia a, Si G The er ab ra o Le n o Ni ne ge M Su ria ad d ag an as c Er ar itr e Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. Zi Be a Cô mb nin te ab d' we Iv o An ire go Et la Road Asset Value as Percentage of the Potential Maximum hi Ce o nt Lib pia ra er lA fri Ch ia ca a n d Bu Rep ru . n Ke di Ta ny nz a M an oz ia am Ma bi li Rw qu an e d Ni a Na ge So m r ut ib h ia Bo Afr i M tsw ca au an rit a a M nia al aw Sw Gha i Bu az na rk ilan in d a M Fa au so rit iu s Roads: The Burden of Maintenance 35 Figure 2.10 Correlation between Percentage of Main and Rural Roads in Good Condition 100 percentage of rural network in good condition 80 60 R2 = 0.2159 40 20 0% 0 20 40 60 80 100 percentage of main network in good condition Source: Calculation by A. Nogales based on data from Gwilliam and others 2009. Figure 2.11 Estimated RAI from Current Network and Percentage of Expansion Needed to Reach 100 Percent RAI 45 600 40 500 35 percentage of expansion 30 400 potential RAI (%) 25 300 20 15 200 10 100 5 0 0 Ni ia Su er Za dan M Ni bia ag ria ut mi r rk Af a aF a oz C aso bi ad Gh ue nz a ne a M gal Ca ga wi er a De Be n m nin te Ken p. d' ya a e so a o So Na sca Bu h bi in ric Ta an Se ani m nd Le nd Rw oir th oo op e g U la ad ge am h q .R m Iv a a hi Et Cô o, M ng Co potential RAI percentage expansion of classified network Source: Tabulation by A. Nogales based on data from Gwilliam and others 2009. Note: RAI = Rural Accessibility Index. 36 Africa's Transport Infrastructure target for the RAI. When these additional kilometers are expressed as a percentage of the current classified network, the results are striking (figure 2.12). Even in the best cases, the classified road network would need to grow in length by around 50 percent, and in most cases it would need to double or even triple in length. Madagascar, evidently an outlier, would need to increase the length of its current classified road network sixfold to attain 100 percent rural accessibility. Isolated rural areas may not be able to realize their full agricultural potential. Hence, another way of assessing the rural network is to look at the extent to which it provides adequate access to high-value agricultural land. For exploration of this effect, estimates by the United Nations Food and Agriculture Organization were compared with actual and potential crop production, using GIS maps. The ratio, indicating the extent of real- ized agricultural potential, was then plotted against the degree of remote- ness (figure 2.12). For most countries, exploitation of potential (for many crops such as cotton, maize, and coffee) was highest in zones between two and five hours' travel time from the nearest large town. Beyond this zone, the ratio of actual production to potential dropped off sharply (the startling out- lier in this graph being Namibia). The reason the highest production is not closer to the towns is that in areas close to towns, agricultural production is Figure 2.12 Accessibility and Agricultural Production 90 80 70 60 percent 50 40 30 20 10 0 greater well connected remote very remote urban connected high actual or potential ratio by access class Benin Cameroon Côte d'Ivoire Kenya Lesotho Madagascar Malawi Namibia South Africa Tanzania Uganda Zambia Source: Carruthers, Krishnamani, and Murray 2009. Roads: The Burden of Maintenance 37 either limited to, or concentrated on, food crops not included in the survey. Lack of accessibility thus appears to be limiting the exploitation of agri- cultural potential in poorer countries with less-dense road networks. This concept is further developed and applied in chapter 7. Road Safety: An Urgent Problem Road safety is a very serious problem in most African countries. Road acci- dent statistics, even those for fatalities, are difficult to obtain, and recent studies have depended on extrapolations from recorded death numbers (see appendix 2j). Nevertheless, it is believed that Africa has 10 percent of the world's road fatalities with only 4 percent of the world's vehicle fleet. In the early years of the new millennium, nearly 3,000 people per year were killed on Kenyan roads. This translates to approximately 68 deaths per 1,000 registered vehicles, which is 30 to 40 times greater than in highly motorized countries. Road traffic crashes are the third-leading cause of death after malaria and HIV/AIDS and present a major public health problem in terms of morbidity, disability, and associated health care costs. The gravity of the road safety problem has now been recognized by governments. At a Pan-African Road Safety Conference, held in Accra in February 2007, government delegates resolved in a joint declaration to make road safety a national health and transport priority and to seek funding for a set of positive actions. These included, among other things, strengthening prehospital and emergency services; mainstreaming safety considerations in road programs; collecting reliable road accident statis- tics; and enacting and implementing national legislation to counter driv- ing under the influence of alcohol, speeding, not using helmets, driving unsafe vehicles, and using mobile phones when driving. Some countries have already taken action: Ghana has established a National Road Safety Council, with subsidiary regional bodies; Uganda has established a Road Safety Education Program; and the South African province of KwaZulu- Natal has launched a comprehensive road safety campaign. Yet in the many nations without an institution dedicated to this issue, it seems that road safety will continue to be ignored. Institutions: Ongoing Reforms Over the past decade, most countries in Africa have followed a consistent path of institutional reform in the road sector. Most countries have a formal transport policy statement, and many have a long-term investment pro- gram. More than 80 percent of the countries studied have adopted formal 38 Africa's Transport Infrastructure sector policies, although most of these policies have not been reviewed in the past five years. Just over 60 percent of countries have a long-term road investment program, in most cases instituted only recently. Such pro- grams, however, depend heavily on foreign aid and cheaply borrowed finance rather than on dedicated and reliable domestic income streams. The central focus of road sector reforms over the past decade has been institutional reform to improve the availability of funds for road mainte- nance and the capacity to execute public works (appendix 2k). Through initiatives such as the SSATP, country governments and development partners have largely come to agree on the establishment of or increase in road funds to provide ring-fenced revenues for road maintenance based on a user charge implemented through fuel levies. A review of the per- formance of second-generation road funds in Africa (Benmaamar 2006) found that while they were steadily improving, their effectiveness was impeded by the inefficiency with which resources were used by the implementing agencies. A second area of action has therefore concerned the establishment of independent road agencies with strong capabilities for the execution of public works (Pinard 2009). These reforms have implications for the line transport ministries, the functions of which should shift from execution to overall supervision. Other institutions of importance include the rural administrations, which are responsible for at least the classified part of the rural road networks, often without any reliable source of funding. The regional economic com- munities have a lesser role--they are primarily concerned with coordinat- ing country actions related to both infrastructure and operations in the transit corridors. Second-Generation Road Funds: Getting Results The aim of establishing second-generation road funds is to improve the condition of the road stock by better funding and more professional man- agement of road maintenance. The philosophy is that road users would be willing to pay increased charges for road use if they were assured that the funds generated would be used for improved maintenance. Eighty per- cent of the initial sample of 24 countries have already introduced road funds, and others are in the process of doing so. Seven design features characterize a "good" second-generation road fund. First, it is important to establish a strong legal basis for road fund operations as a protection against ad hoc political interference. Such a basis ideally entails a concise enabling law supported by published regulations specifying how the fund is to be managed. Fifteen (60 percent) of the Roads: The Burden of Maintenance 39 sample countries with road funds have such founding legislation; the rest have relied on decrees. But the quality of the legislation is not uniformly high: a review in 2004 concluded that many of the funds were poorly designed, with limited administrative or financial autonomy and inade- quate auditing provisions. Second, the functions of road funding and road service provisions should be separated, with both undertaken by autonomous agencies. The creation of autonomous road agencies for public works execution has generally lagged behind that of road funds. At present, about 65 percent of the countries with quasi-independent road funds also have an inde- pendent implementation agency, with implementation undertaken in other countries by departments of the relevant central ministry. Third, the fund should be financed by user charges entirely independ- ent of any fuel taxes that may meet general revenue purposes. About 80 percent of the sample countries have established road user charges, typically in the form of fuel levies. But in many cases, the fuel levy is set well below the level needed to cover the maintenance costs arising from wear and tear of the network by road users, let alone contribute to fund- ing the rehabilitation backlog (figure 2.13). Figure 2.13 Fuel Levy Relative to Optimal Requirements for Maintenance and Rehabilitation 150 fuel levy relative to optimal 125 benchmarks (%) 100 75 50 25 0 r da o r i a a n ia a e m a n a aw ge ca bi i ny i an qu th ni oo ib op an an as m Be Ni so al Ke m Gh bi er nz hi Rw Za ag M Na Le am Et Ta ad Ca oz M M percentage of optimal maintenance percentage of optimal maintenance plus rehabilitation Sources: SSATP 2007; Briceño-Garmendia, Smits, and Foster 2009. 40 Africa's Transport Infrastructure In practice, the fuel levy varies widely across countries, from 3 cents per liter in Lesotho to 16 cents per liter in Tanzania (figure 2.14, panel a). Moreover, the fuel levy collection rate also varies substantially. Four coun- tries (Ghana, Niger, Rwanda, and Tanzania) collect only about half of the fuel levy revenue that should go to the road fund. The problems responsi- ble for the shortfall range from widespread tax evasion in Tanzania to administrative problems in the transfer of revenues from the collection agency to the road fund in Rwanda (figure 2.14, panel b). In some cases (notably Ethiopia and Madagascar), the ratio of actual funding to estimated Figure 2.14 Average Fuel Levy across Countries with Second-Generation Road Funds, 2007 a. Fuel levy b. Implicit fuel levy collection ratio Tanzania Ethiopia Namibia Madagascar Kenya Benin Mozambique Malawi Cameroon Kenya Ethiopia Zambia Malawi Cameroon Ghana Namibia Niger Lesotho Côte d'Ivoire Mozambique Benin Zambia Niger Rwanda Rwanda Madagascar Tanzania Lesotho Ghana 0 2 4 6 8 10 12 14 16 0 50 100 150 200 fuel levy (US cents per liter) implicit fuel levy collection ratio (%) Source: SSATP 2007; Briceño-Garmendia, Smits, and Foster 2009. Roads: The Burden of Maintenance 41 fuel levy revenue is well above 100 percent, indicating substantial central government transfers to the road fund. In several cases (Benin, Côte d'Ivoire, Ethiopia, Gabon, and Zambia), the fund is dependent on budget allocations for more than 75 percent of its resources. Few road boards have effective power to adjust fuel levies in line with changing mainte- nance requirements, because of residual controls by the ministry of finance over the level of fuel levies. Fourth, road user charges should be transferred directly to the road fund without passing through the government budget. Channeling of fuel levy revenues through the budget increases the risk that the rev- enues may be diverted to finance other public expenditures. Just over 50 percent of the sample countries with road funds successfully channel a high percentage (that is, at least 75 percent) of their fuel levy revenues directly to the road fund. In other cases, direct channeling covers a very low proportion of fuel levy revenues (less than 25 percent) or none at all, making the resource base for road funds much more vulnerable to diversion. Fifth, user representation on the road fund board helps to strengthen accountability. It also allows users to make direct trade-offs between the level of user charges and the quality of the road network. With the excep- tion of Malawi, all the countries with road funds have established inde- pendent road fund boards. But half of the boards still have a majority of government representation, with the chairman and executive secretaries usually being political appointees. Sixth, to reduce discretion in fund allocation, clear and explicit rules for the allocation of funds to different types of expenditures are needed. About 60 percent of the road funds surveyed have established percent- age allocations for dividing funds among different portions of the road network, although the chosen allocations differ substantially across countries (figure 2.15). On average, about 60 percent of the resources go to the main road network. Around half of the countries are allocat- ing at least 20 percent of the road fund resources to the rural road net- work. Overhead typically accounts for no more than 6 percent of road fund revenues, even though the number of professional staff members employed varies widely, from only 6 in Niger to 48 in Kenya (the large size of the staff in Djibouti is due to the fact that employees are also involved in collecting transit fees). Seventh, independent technical and financial auditing and public reporting of the road fund activities help to strengthen accountability (Heggie and Vickers 1998). About 80 percent of the countries with road 42 Africa's Transport Infrastructure Figure 2.15 Overview of Road Fund Allocation Rules 100 90 percentage of road fund allocation 80 70 60 50 40 30 20 10 0 am da e a Za i Na ia ia a r Et n Ta ia ad nia r d' d n re aw ge ca an ny qu oo ni a b ib op oi an Ch as a m Be Ni al Ke Gh m bi er nz Iv hi Rw ag M m Ca te oz M Cô M main roads rural roads urban roads overheads other Source: SSATP 2007. funds report that auditing procedures are in place. In most cases, these cover both technical and financial auditing and take place on an annual basis. But the quality of these audit processes is dubious in some coun- tries. The prevalence of financial auditing is somewhat higher than that of technical auditing. The SSATP Road Maintenance Initiative policy matrix shows that the prevalence of each of these criteria differs significantly (figure 2.16, panel a) and that the overall score for road fund design also varies widely across countries (figure 2.16, panel b). On average, the road funds in the sample countries meet 65 percent of the defining criteria for second- generation road funds. There is a broad performance range, from coun- tries such as Tanzania, Namibia, and Kenya that fulfill 100 percent of the criteria to countries such as Benin and Burkina Faso that appear to cap- ture well below 50 percent of them. Another important achievement of road funds has been to stabilize, increase, and improve the predictability of maintenance expenditures. The volatility of road fund expenditures (measured by calculating the standard deviation around the trend line) was shown to be only half that of expenditures arising from external funding and one-third that of central Roads: The Burden of Maintenance 43 Figure 2.16 Evaluation of Road Fund Reforms a. Prevalence of second-generation characteristics b. Scores on overall performance index Tanzania independent auditing Namibia Kenya Rwanda* road user charges Malawi Madagascar Ghana revenue allocation rules Ethiopia Chad separation of Zambia functions Niger* Mozambique clear legal Cameroon basis Lesotho Côte d'Ivoire direct transfer Benin South Africa Senegal user representation on Cape Verde board Burkina Faso 0 20 40 60 80 100 0 20 40 60 80 100 percentage of countries score (%) Source: SSATP 2007. * = based on incomplete information. government allocations in time-series data over the period 2001­05. Moreover, the volatility of road fund expenditures appears to be lower in countries that have made efforts to ensure the independence of their road funds and have increased the proportion of revenues channeled directly into the funds. Road Agency Performance: Lagging Behind In many countries, the weaknesses of traditional maintenance by force account are now well recognized. It was initially thought that the prob- lems associated with timely and cost-effective implementation of public works contracts could be solved by reforming and restructuring the road departments housed in the line ministries for the sector. But restructur- ing of road departments has not had the expected beneficial impact on road project implementation, in part because too many constraints still 44 Africa's Transport Infrastructure prevent the full use of existing technical capacity. One important con- straint has been staff skills and leadership. The economic growth in the region over the past decade has increased the demand for engineers in the private sector, which has attracted better-qualified staff with higher salaries. But road departments typically lack employees with the skills needed to review the design, costs, and work under various contracts. This situation prolongs the contracting process and may be a reason for the recent escalation in the unit costs of road construction. For those reasons, current thinking focuses on moving the responsibil- ity for managing implementation out of the traditional civil service struc- ture into an independent agency. About two-thirds of the sample of countries in the SSATP survey have already established some sort of inde- pendent, commercialized roads agency, and a number of others are in the process of doing so. But only a third of these have private sector represen- tation on their boards. Levels of autonomy vary from full responsibility for road network management to limited responsibility for the execution of road maintenance programs defined by the roads department or min- istry of roads. A recent study by Pinard (2009) identified two quite different institu- tional forms, namely (i) a roads agency, which, though a legal entity, is not independent of its ministry, and (ii) a roads authority, which is essentially an independent legal identity. Pinard compared the two types in terms of their institutional characteristics (a combination of their legal foundation, composition, powers, and processes) and their performance. He found that paved road conditions were better under the independent authori- ties, a trend he attributed largely to the authorities' greater autonomy. But even with an authority, problems remain. Most road authorities are still not able to pay market-based wages--staff salaries are typically 60 to 80 percent of those for similar jobs in the private sector--making staff recruitment and retention difficult. Many road authorities fulfill aspects of the "supplier" function and undertake varying amounts of noncore activities, which reduces their focus on managing performance. Many are unable to operate their road asset management systems to produce reli- able outputs in terms of optimal network strategies and programs. This deficiency suggests that more aspects of data collection and system oper- ation should be contracted out to competent consultants, though both the number and the capacity of local consultants and contractors are lim- ited in a number of countries. Hence, although road authorities are improving governance, attracting skilled staff and ensuring continuous collection of reliable data remain a challenge. Roads: The Burden of Maintenance 45 In about half of the sample countries, more than 80 percent of main- tenance work was contracted out. This approach was strongly, though not exclusively, associated with the presence of a road agency in the country. Contractors were typically paid directly by the road fund, usually in fewer than 30 days for undisputed bills; however, in Burundi, Ghana, and Kenya, the average payment time was 90 days in 2006. Improved contract management and disbursement arrangements have resulted in a 10 to 20 percent reduction in road maintenance costs per kilometer in Ethiopia, Ghana, and Zambia. An associated development has been the establishment of performance- based contracts with the private sector for road maintenance on a longer- term, multiyear basis. Such contracts have been made possible by the greater security of road maintenance revenues resulting from the establish- ment of second-generation road funds. They are advantageous in that they provide a strong incentive for contractors to undertake effective mainte- nance activities and to reduce expenditure uncertainties for the road fund. But it is already clear that the benefits of shifting road work from force account to private performance-based contracting depend on the exis- tence of an efficient and competitive road-contracting industry, a transpar- ent process of selecting contractors, and an ability to negotiate and manage the contracts effectively (Stankevich, Qureshi, and Queiroz 2005). In par- ticular, ensuring that trucks are not overloaded is important for the imple- mentation of long-term performance contracts. Rural Road Administration: The Orphaned Sector In many countries, the responsibility for the rural segment of the network is devolved to the local level (Malmberg Calvo 1998). There are two dis- tinct administrative categories of rural transport infrastructure, namely, local government roads and community roads and tracks. The former are designated as the responsibility of the appropriate local government unit; the latter have no formal owner. While community facilities may have been built by nongovernmental organizations or even by foreign-aid agen- cies, they tend to be neglected if they have not been formally assigned to any agency for their subsequent maintenance. For example, a Zambian nongovernmental organization built 1,000 km of roads during the early 1990s as part of a drought-relief effort, but the roads have deteriorated badly because no institution is legally responsible for them. Sources for financing local government roads are usually very limited. Local governments mobilize only modest revenues of their own, the main sources often being market and business taxes. Intergovernmental transfers 46 Africa's Transport Infrastructure are usually the main source of domestic funding for local governments. Three main problems result from relying on the central budget for fund- ing maintenance of rural roads. First, throughout most of Africa, less than 5 percent of aggregate public sector revenue is generally made available to rural governments. Second, general budgets rarely allocate adequate funds for maintaining main roads, much less rural roads. Third, capital and recurrent allocations to local governments are usually not fungible, and the allocation for recurrent expenditures may barely cover the salary expenditures of the local rural road unit. Moreover, such transfers are dic- tated by the budget cycle, so they are unlikely to provide an adequate and timely source of funding. Adequate and steady funding of local govern- ment road maintenance is more likely to come from a dedicated road fund, as long as the road fund law expressly states that the fund accepts responsibility for local roads. Local government networks tend to be small, often too small to attract the interest of competent consulting firms to manage their maintenance. In Madagascar, the average network size for a local government is 140 km; in Cameroon and Nigeria, 180 km; and in Tanzania and Zambia, 280 km. But a network size of 500 to 2,000 km is usually required to justify employing an engineer. Joint services committees of local authorities may achieve economies of scale in procurement, but they usually require sub- stantial technical assistance from central ministries or from regional offices of a main-roads authority. In countries with an autonomous authority over the main roads, local governments may contract this authority to manage roads on their behalf or to assist with planning and procurement. Private sector capacity and capabilities can also be mobilized by contracting out physical works or even key management functions to local consultants. Specialized contract management agencies known as AGETIPs are com- mon in Francophone Africa--for example, in Madagascar, Mali, Niger, and Senegal. These agencies manage and use private consultants and contrac- tors on behalf of the public authority and perform all the necessary func- tions of contract preparation, implementation, and supervision. Some countries centralize the technical responsibility for rural roads. This practice has the advantage of enabling better technical support, but because the central authority often operates independent of the local government structure, it is usually poorly connected to local needs and developments. In principle, a central coordinating unit for local government roads should be able to perform as well as a central government rural-roads department. In practice, however, such coordinating units for local government roads are weak, as they were in Tanzania and Zambia in the late 1990s. Roads: The Burden of Maintenance 47 A comprehensive study of road administrations in three countries (Mexico, Uganda, and Zambia) has found that decentralization has yielded few of the expected advantages (Robinson and Stiedl 2001). Several factors contributed to this outcome: lack of local government powers to exercise political influence, insufficient financial resources, lack of management capability, and lack of accountability mechanisms at the local level. It was concluded that countries contemplating decentraliza- tion were most likely to benefit from the "devolved and delegated" model (local government is the owner, with a parastatal or private sector admin- istrator working under contract) or the centralized road fund model. These options were not mutually exclusive. For example, a joint services committee may use private consultants, hired through a contract manage- ment agency. The best option for managing local government roads depends on many local factors, including the size of the authority, the nature of the network for which it is responsible, and the competence of the sector or higher-level public authority units. Community infrastructure faces particular problems. Community con- tributions in cash and in kind (usually labor) are suitable primarily for community roads and paths. But contributions in kind may produce rel- atively inefficient labor, making other sources of money necessary. Strategically designed cost-sharing arrangements for local government roads and community roads may stimulate resource mobilization at all levels and increase the proportion of the networks receiving regular main- tenance. Well-structured donor financing through rural road projects or through social funds or rural infrastructure funds can assist investment in community-level infrastructure as well. Cost-sharing arrangements may also be effective in maintaining community roads. Many local authorities in Africa have more roads to maintain than they can afford. Achieving effective community management is often impeded by lack of technical know-how. Communities in Africa therefore need technical advice (for instance, on road design and standards, appropriate materials, and work planning) and managerial advice (in areas such as financial accounting, contract management, and procurement) so that they can effectively per- form the responsibilities that come with ownership of roads and paths. Road Spending: A Problem of Execution It is important to recognize the distinction between road funding (the process of budget allocation) and road spending (the actual execution of the budget). These can differ substantially, either because delays in the 48 Africa's Transport Infrastructure budget process leave too little time for execution within the fiscal year or because of lack of capacity in the road construction sector. The percentage of national income actually spent in the road sector in the initial AICD sample countries, taking into account all budget and extrabudgetary channels (such as road funds), has been estimated in the AICD fiscal costs study (Briceño-Garmendia, Smits, and Foster 2009). This analysis shows that, on average, the sample countries devote 1.8 per- cent of their GDP to the road sector (figure 2.17). This is within the 1 to 2 percent range of expenditures found in those countries around the world with already well-developed infrastructure and GDP growth rates of 2 to 3 percent, but it is below the levels found in a number of fast- growing countries that have made intensive efforts to upgrade transport Figure 2.17 Average Annual Expenditures on Road Transport by Country, 2001­05 4.5 35 4.0 30 3.5 25 3.0 percentage of GDP US$ per capita 2.5 20 2.0 15 1.5 10 1.0 5 0.5 0.0 0 te Be n D' nin Ke ire an a Ni a Ni r G ria nz a n a an l ad Ch a ag ad m r oz hi a bi ia M ue i V ca m e so a o Ug ega aw ge Za sca Rw ny d Ta han Se ani d M Et bi Le ibi Na erd th oo am op pe fri ge q o al Iv er a Ca A m h ut Ca So M Cô as % GDP US$ per capita Source: Briceño-Garmendia, Smits, and Foster 2009. Roads: The Burden of Maintenance 49 infrastructure. For example, Brazil, India, the Republic of Korea, and the former Soviet Union all invested between 2 to 3 percent of their GDP in transport infrastructure during the 1980s, while between 2000 and 2002, Korea, Malaysia, and Thailand were investing 1.7 percent to 1.9 percent of their GDP and achieving GDP growth rates between 4 and 6 percent. Thus, while the African expenditure effort is not far below that achieved in many developing countries, it does fall short of that associated with very high rates of economic growth. Average spending on roads in Africa varies from less than 1 percent of the GDP in South Africa to almost 4 percent in Malawi. The highest income shares are found in the poorest countries. For the middle-income countries in the sample, spending tends to be clustered around 1 percent of the GDP, but in all countries, the absolute values remain small, at around $7 per capita per year for low-income countries and $22 per capita per year for middle-income countries. Virtually all road expenditure is in the public sector. Even in the future, the relatively modest traffic volumes projected for most corridors mean that the scope for privately funded toll roads is limited (box 2.1). The same aggregate information on road expenditure can also be expressed as a rate per kilometer of main road network (again, composed of roads managed by the central government). In most countries, the main network comprises the primary and secondary networks, but in a few cases it is limited to the primary network. On average, the African coun- tries investigated in the fiscal costs study spent just over $9,000 per kilo- meter of main road (table 2.4). But spending levels in low-income countries are more than 50 percent higher per kilometer than spending levels in the middle-income countries, with resource-rich, low-income countries spending slightly more than aid-dependent ones. Landlocked countries and islands spend substantially more per kilometer than what is spent by coastal nations, which may be attributable to higher costs of importing materials and services. Countries with rolling terrain and humid conditions, which tend to accelerate road deterioration, show somewhat higher levels of spending than countries with flat terrain and arid conditions. Some of the observed outcomes are paradoxical. For example, countries with road agencies seem to spend substantially less than those without them, irrespective of whether they have road funds, and those countries with low fuel levies actually spend substantially more on roads than those with no or high fuel levies. To resolve those paradoxes, one must look further into the composition of the spending and the sources from which it is financed. 50 Africa's Transport Infrastructure Box 2.1 Road Concessions in Africa Only 10 African toll road projects are recorded for the years since 1990 in the World Bank's Private Participation in Infrastructure (PPI) database. These include eight projects in South Africa alone, one an international road corridor connect- ing South Africa to Mozambique. The other two projects involve the construction of bridges over the Abidjan Lagoon in Côte d'Ivoire and the Limpopo River in Zimbabwe. The projects are quite evenly divided among greenfield projects, con- cession contracts, and lease contracts. Overall, only 1,600 km of Africa's total classified road network of 1.2 million km have been contracted out to the private sector under a medium- or long-term man- agement arrangement. The total cumulative private sector investment committed under these projects amounts to $1.6 billion, barely 20 percent of the estimated annual investment needed in Africa's road sector ($7.6 billion). The potential for toll road concessions in Africa remains limited because of the relatively low traffic flows in the region. Based on the AICD sample of countries, only 8 percent of the region's road network (that is, less than 9,000 km) has traffic levels in excess of 10,000 vehicles per day, which is the threshold to make toll road concessions economically viable. Some 86 percent of those viable kilometers are concentrated in South Africa, and a further 8 percent are in Nigeria. A number of other countries have up to 100 km of paved road at this traffic level, but many oth- ers do not reach this level of traffic in any segment of their paved road network. Source: Author, based on World Bank 2008a. Capital Investment: More than a Fair Share of Spending A strong capital bias is evident in road sector spending. Analysis of African road needs suggests that about half of road sector spending should go to capital projects and the other half to maintenance of existing assets. In reality, about two-thirds of spending is allocated to capital projects in the 19 countries studied (figure 2.18). The bias is most pronounced in low-income countries, those with chal- lenging geographical environments, and those without road funds or fuel levies. There is a very striking difference between the middle-income countries, which devote only 25 percent of their road spending to capital projects, and the low-income countries, which devote around 70 percent to capital projects (table 2.5). Roads: The Burden of Maintenance 51 Table 2.4 Average Annual Expenditures per Kilometer of Main Road by Country Category, 2001­05 Average annual Average annual Country spending on roads Country spending on roads characteristics (US$ per km) characteristics (US$ per km) Income level Institutions Middle-income 6,050 Road fund and road 7,112 Low-income, aid- 8,823 agency dependent Road fund only 9,793 Low-income, resource- 9,551 Road agency onlya 6,053 rich Financing Geography Low fuel levy 9,458 Coastal 7,014 High fuel levy 8,117 Island 13,302 No fuel levya 7,153 Landlocked 9,984 Topography Flat and arid 7,977 Rolling and humid 9,518 Source: Briceño-Garmendia, Smits, and Foster 2009. a. South Africa is excluded from this group. Figure 2.18 Percentage of Road Spending Allocated to Capital Projects 100 percentage of total spending (%) 80 60 40 20 0 Ke a nz a Be a M n Za wi m ia M G on bi a Le ue o Rw ger Et nda M ga a ag a Ni ar Cô Sen ia d' al re ad ric ny i am n U pi ad nd th ni an Ca mb r te eg c oi a oz ha ge Ch q o as o Ni so al Af a lv er hi h Ta ut So Source: Briceño-Garmendia, Smits, and Foster 2009. To some extent, this difference may reflect the fact that the low- income countries are still developing transport networks, whereas the middle-income countries have already established their basic transport platform and can devote themselves predominantly to maintenance. 52 Africa's Transport Infrastructure Table 2.5 Percentage of Road Spending Allocated to Capital Projects, by Country Category Spending for Spending for capital projects capital projects Country (as % of all road Country (as % of all road characteristics spending) characteristics spending) Income level Institutions Middle-income 25 Road fund and road 58 Low-income, aid- 68 agency dependent Road fund only 64 Low-income, 77 Road agency only a 86 resource-rich Financing Geography Low fuel levy 72 Coastal 53 High fuel levy 45 Island 85 No fuel levy a 85 Landlocked 74 Topography Flat and dry 58 Rolling and humid 72 Source: Briceño-Garmendia, Smits, and Foster 2009. a. South Africa is excluded from this group. Countries facing difficult geographic and topographic conditions also show evidence of a stronger bias toward capital expenditure. Countries with road funds show a lower degree of capital bias than those without, irrespective of whether they have independent road agencies or not. Countries with high fuel levies show no evidence of capital bias. Even these relatively high levels of capital expenditure understate the true extent of capital bias in road spending. The reason is that, on average, only around 70 percent of budgeted capital spending is actually executed within the corresponding budgetary cycle due to weaknesses and delays in the public procurement process. These delays prevent contracts from being awarded and completed within the 12-month budget cycle (figure 2.19). There are substantial and systematic variations in budget execution across countries and country groupings. Budget execution ranges from 25 percent in Benin to over 100 percent in South Africa (table 2.6). There are also systematic differences across country categories. Middle-income countries perform substantially better than low-income countries, and countries with road funds and fuel levies perform substan- tially better than those without. There is also a striking difference in favor of countries with rolling, humid terrain relative to those facing flat, arid Roads: The Burden of Maintenance 53 Figure 2.19 Capital Budget Execution Ratios 150 capital budget execution ratio (%) 125 100 75 50 25 0 n ia a ria ad i na a da ia n ad nia r a aw ca ny bi ric ni oo ib op a ge Ch an as a m Be al Af Ke m Gh er nz hi Ug Ni Za ag M Na m h Et Ta ut Ca So M Source: Briceño-Garmendia, Smits, and Foster 2009. Table 2.6 Capital Budget Execution Ratios, by Country Category Percentage Percentage Macroeconomy Institutions Middle-income 83 Road fund and road agency 66 Low-income, aid-dependent 67 Road fund only 64 Low-income, resource-rich 61 Road agency only a 43 Geography Financing Coastal 64 Low fuel levy 65 Island 92 High fuel levy 62 Landlocked 71 No fuel levy a 59 Topography Flat and dry 63 Rolling and humid 78 Source: Briceño-Garmendia, Smits, and Foster 2009. a. South Africa is excluded from this group. conditions, perhaps indicating the greater urgency of road works in the former setting. Above average capital expenditure on roads may be justified by large rehabilitation backlogs. Using the RONET model, it is possible to pro- duce detailed estimates of the requirements for rehabilitating each coun- try's road network, taking into account the current distribution of network conditions and working toward a target of clearing the current rehabilitation backlog within a reasonable period of time. On that basis, the rehabilitation requirements can be compared with the current levels 54 Africa's Transport Infrastructure of capital expenditure to determine whether these are high enough to eliminate the rehabilitation backlog within a five-year period (figure 2.20). Negative numbers indicate that the current levels of expenditures are not sufficient to eliminate the backlog. It is important to note that this calculation is only illustrative and is based on the assumption that the entire capital budget is devoted to network rehabilitation. While rehabilitation usually dominates capital spending, some upgrad- ing of road categories and addition of new roads does occur. Although the available data do not make it possible to know the exact split, the calcu- lation in table 2.7 is helpful in indicating whether current levels of capi- tal expenditure would be high enough to address the rehabilitation problem if they were fully allocated to rehabilitation works. In fact, only in half the countries is capital spending high enough to reasonably address rehabilitation backlogs. In the other half, capital spending has fallen well below what is needed to clear rehabilitation backlogs. Chad and Ethiopia stand out as countries undergoing very large road investment programs, including major works to upgrade road categories and extend the reach of the networks. In these cases, spending is two to three times the level needed to clear rehabilitation backlogs. Countries with both a road fund and a road agency seem to show the highest margin of capital spending Figure 2.20 Deviation of Capital Expenditure from Expenditure Required to Meet Rehabilitation Requirements within a Five-Year Period 400 350 deviation from requirement (%) 300 250 200 150 100 50 0 ­50 ­100 Be a n Ni r M ia er i Ug on Za da Le bia M wa o M ag a bi r Se ue d' al Gh re Ta ana ia Et ad ia Ca aw ge am ca ny ad nd R th ni r eg an op oi ge an Ch q o oz as m Ni so al Ke lv nz Cô n hi m te rehabilitation rehabilitation adjusted for capital execution Source: Briceño-Garmendia, Smits, and Foster 2009. Roads: The Burden of Maintenance 55 Table 2.7 Capital Expenditure as Percentage of Rehabilitation Needs, by Country Category Percentage deviation of actual annual total capital expenditure from expenditure necessary to eliminate accumulated rehabilitation needs over a period of five years Percentage Percentage Macroeconomy Institutions Middle-income ­6 Road fund and road agency 60 Low-income, aid-dependent ­3 Road fund only ­19 Low-income, resource-rich 22 Road agency onlya ­27 Geography Financing Coastal ­21 Low fuel levy ­5 Island ­4 High fuel levy 24 Landlocked 30 No fuel levya ­28 Topography Flat and arid ­7 Rolling and humid 13 Source: Briceño-Garmendia, Smits, and Foster 2009. a. South Africa is excluded from this group. over rehabilitation requirements. Resource-rich, low-income countries, landlocked countries, and countries with high fuel levies also tend to show capital spending that is somewhat higher than rehabilitation needs. Failure to execute the budget is common. Budgeted capital spend- ing is typically 40 percent higher than what countries actually succeed in spending. Hence if rehabilitation requirements are compared to an estimate of budgeted (versus actual) capital spending, the funding sit- uation looks somewhat more positive, with the percentage of countries able to meet their rehabilitation requirements within a reasonable time period increasing from one-half to two-thirds. Thus, improving capital budget execution is an important first step toward clearing rehabilita- tion backlogs. Public investment in roads is highly dependent on flows of aid, which can be volatile. It is not always possible to trace with precision the items on the public investment budget that are financed by official develop- ment assistance. The limited evidence available indicates a heavy depend- ence on foreign funding, which ranges from just over 50 percent in Senegal to almost 90 percent in Rwanda (figure 2.21). The volatility of official development assistance flows contributes to the volatility of pub- lic investment in the sector. Thus, the very high ratios of road investment to GDP in Chad in 2003­05, in Tanzania in 2000, and in Madagascar in 2004­05 were all associated with short-lived surges in aid. 56 Africa's Transport Infrastructure Figure 2.21 Foreign Funding as Percentage of Capital Spending 100 75 percent 50 25 0 e on ia l a a n ad r ga qu d d ge ni an an an Ch o ne Be bi Ni er nz Ug Rw Se am m Ta Ca oz M Source: Briceño-Garmendia, Smits, and Foster 2009. Higher construction standards result in slower deterioration rates and lower annual maintenance costs. They also counteract, to some extent, the adverse effects of vehicle overloading, which is rife. As noted above, the AICD data already show a negative correlation between capital and maintenance expenditures as well as the underfunding of maintenance expenditures. Given the reduction in operation costs per tonne-km as vehicle loadings increase, it may be sensible for countries to jointly recon- sider their policies on construction standards and vehicle axle weights. Maintenance Expenditures: Squeezed There appears to be a trade-off between levels of capital expenditure and levels of maintenance expenditure, shown by the large negative correla- tion (­0.33) between the level of maintenance expenditure per kilome- ter of the main network and the level of capital expenditure per kilometer (figure 2.22). This can be plausibly explained. On the one hand, countries that spend too little on maintenance will end up with larger rehabilitation liabilities, often resulting in the need for emergency works to restore the functionality of critical infrastructure. On the other hand, countries with large investment programs may have fewer resources left over to address road maintenance needs. The latter scenario is worrisome because if high capital spending comes at the expense of lower maintenance expenditure, then the condition of the network will only deteriorate further over time. Roads: The Burden of Maintenance 57 Figure 2.22 Relationship between Capital Spending and Maintenance Spending per Kilometer of Main Network 5,000 4,500 4,000 maintenance spending 3,500 (US$ per km) 3,000 2,500 2,000 1,500 1,000 500 0 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 16,000 capital spending (US$ per km) Source: Briceño-Garmendia, Smits, and Foster 2009. There is huge variation in maintenance expenditure efforts, both across countries and across rural and main road networks. For the main road net- works, the range extends from barely $200 per kilometer in Chad to over $6,000 per kilometer in Zambia. For the rural road networks, the range extends from barely $20 per kilometer in Chad to more than $3,000 per kilometer in Lesotho (figure 2.23). On average, countries are spending $1,100 per kilometer on rural networks and about double that amount, or $2,200 per kilometer, on the main networks. Indeed, some countries are spending more on maintenance per kilometer for their rural networks than other countries are spending on maintenance per kilometer for their main road networks--as is the case with Tanzania and Madagascar. Overall, the correlation between maintenance efforts on main networks and those on rural networks is positive and high (0.36) across countries, which is to say that countries that tend to spend larger amounts on main network maintenance also tend to spend larger amounts on rural network maintenance and vice versa. For a comparison of countries, two different standards were hypothe- sized (appendix 2l). The "custom" standard assumes that all primary roads are kept in good condition and secondary roads in fair condition, with other roads allowed to be in poor condition. The "optimal" standard links the standard to be achieved to the traffic volume on any network, with the total maintenance budget optimized to reduce total system operating and maintenance costs. 58 Africa's Transport Infrastructure Figure 2.23 Average Maintenance Spending across Different Parts of the Network (US$ per km) a. Main network b. Rural network Zambia Lesotho South Africa Tanzania Namibia Zambia Tanzania Ethiopia Benin Namibia Kenya Niger Nigeria Cameroon Kenya Ethiopia Mozambique Ghana Ghana Mozambique Rwanda Malawi Malawi Uganda Benin Madagascar Cameroon Niger Madagascar Rwanda Lesotho Uganda Chad Chad 0 1,000 2,000 3,000 4,000 5,000 6,000 0 1,000 2,000 3,000 4,000 5,000 6,000 maintenance spending (US$ per km) maintenance spending (US$ per km) Source: Briceño-Garmendia, Smits, and Foster 2009. Even more important than the absolute spending is the comparison between spending and requirements. If one uses the RONET, it is possible to produce detailed estimates of the routine and periodic maintenance requirements needed to preserve each country's road network to the cus- tom standard (appendix 2m) or to the optimal standard (appendix 2n), taking into account the current distribution of network conditions. It is important to note that this calculation is based on the assumption that the entire maintenance budget is spent on maintenance works at efficient unit costs. The results of this comparison are shown in figure 2.24 for the custom standard. Appendixes 2o and 2p show the distribution of expen- diture by type of work for these two standards to be achieved. Roads: The Burden of Maintenance 59 Figure 2.24 Deviation of Actual Maintenance Expenditure from That Required to Attain Custom Standard of Maintenance 600 500 maintenance expenditure as % of custom requirements 400 300 200 100 0 ­100 M ire Et car e a Za n Ug ria da d' l am ia ag i Le pia Rw ho Gh a m ia Ni ad r Ke a h a Ta ya Na nin te ga ad aw ge ric d an ut bi qu oo oz an Ca mib n ge an an Ch o t as Cô ene o So m Be Ni so M al Af Iv bi er M nz hi S routine plus periodic maintenance routine maintenance Sources: Briceño-Garmendia, Smits, and Foster 2009; calculation by A. Nogales based on data from Gwilliam and others 2009. This exercise shows that half of the countries are not devoting ade- quate resources to routine and periodic maintenance of the main road networks. In countries such as Chad, Niger, Nigeria, Senegal, and Uganda, maintenance spending comes to less than half the norm requirements. Moreover, around a quarter of the countries are not devoting enough resources to cover even routine maintenance activity. Table 2.8 shows that underspending on maintenance to the custom standard is evident in the low-income countries (particularly the resource- rich countries) and in countries with difficult geographical environments and terrain. Middle-income countries tend to spend substantially above the maintenance norm. Of the six countries not covering even routine maintenance, two were without road funds and levies. Among countries with fuel levies, those with high levies did substantially better than those with low ones. The network preservation costs estimated by the RONET, including both maintenance and rehabilitation for the entire classified network, can be compared to the GDP to gauge their overall affordability at the coun- try level (figure 2.25). The estimated average annual cost of preserving the classified road network lies in the range 0.2 to 4.1 percent of GDP. 60 Africa's Transport Infrastructure Table 2.8 Actual Maintenance Expenditure as Percentage of Expenditure Required for Custom Maintenance Standard, by Country Category Maintenance Maintenance Country spending (as % spending (as % characteristics of requirement) Country characteristics of requirement) Income level Institutions Middle-income 80 Road fund and road ­11 Low-income, ­12 agency aid-dependent Road fund only ­3 Low-income, ­28 Road agency onlya ­69 resource-rich Financing Geography Low fuel levy ­19 Coastal 20 High fuel levy 28 Island ­45 No fuel levya ­69 Landlocked ­24 Topography Flat and arid 12 Rolling and humid ­24 Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Numbers in the table indicate the percentage deviation of actual maintenance expenditure from amount required for custom standard of maintenance. a. South Africa is excluded from this group. Most countries lie in the range of 0.5 to 2.0 percent of GDP per year. Only three countries (Central African Republic, Liberia, and Zimbabwe) are estimated to need expenditures in excess of 2.5 percent of GDP per year. Overall, these numbers do not look very high compared to the data on real historic expenditure reported above. The RONET calculations are based on efficient unit costs, however, and hence are probably an under- estimation of the actual expenditure needs. Road Work Costs: The Toll of Inflation Available data, though limited, indicate that, on average, maintenance costs in Africa, at $2,160 per kilometer, are higher than the worldwide average of $2,024 per kilometer and twice as high as those in South and East Asia. These data suggest that routine maintenance is somewhat less effectively performed in Africa than in other regions (table 2.9). Moreover, there has been a marked increase in unit costs in recent years, large enough to undermine the adequacy of road funding. A recent unit cost study, undertaken as part of the AICD (Gwilliam and others 2009), analyzed data from bills for 115 recently completed donor-funded 61 Figure 2.25 percentage of GDP 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Su d An an go l M G ab a au o rit n So Nig ius ut e r Co h A ia ng fric o, a E Re C thi p. Cô ame opi te ro a d' on Iv oi Ke re Rw ny a a Se nda ne g Ch al Be ad Bu Za nin rk m Source: Calculation by A. Nogales based on data from Gwilliam and others 2009. in bi Bo a Fa a ts s o w Ta an n Sw za a az nia ila M G nd ad h a ag n a as c Ni ar M U ge oz ga r am nd bi a qu e M Gu al in i M e M ala a au w ri i Na tan m ia Co ib ng To ia o, E go D e rit m ea . Le Rep so . Ce Si Bu tho nt er ru ra l A G ra L nd fri am eo i ca bi ne n a, Re Th pu e b Zi Lib lic m e ba ria bw e Aggregate Requirements over a 20-Year Period for Preserving the Road Networks as Percentage of the Current Annual GDP 62 Africa's Transport Infrastructure Table 2.9 Recent Estimates of Unit Costs of Road Maintenance and Rehabilitation World Bank ROCKS databasea AICD unit cost study $ per km Other developing regions Africa Africa Routine maintenance 2,000 2,200 -- Periodic maintenance 43,000 54,000 158,000 Rehabilitation 191,000 162,000 300,000 Sources: World Bank 2008b; calculation by A. Nogales based on data from Gwilliam and others 2009. Note: -- = not available. a. ROCKS = Road Costs Knowledge System road contracts in Africa. The unit costs from this study are two to three times as high as those found in the World Bank road cost database, ROCKS (World Bank 2008b). As a result, a number of donors are find- ing that their road projects experience cost overruns ranging from 20 to 120 percent relative to expectations based on initial engineering designs. Those cost increases, if general, are large enough to seriously affect the adequacy of road sector maintenance and rehabilitation expenditure. A more detailed investigation of cost overruns identifies three explana- tory factors. First, a lack of effective competition--defined as having a price spread of no more than 10 percent among the three lowest bidders-- is strongly associated with the presence of cost overruns in road mainte- nance and rehabilitation contracts. Second, since 2002 and especially since 2005, prices for the basket of items that are key inputs into road construction (such as bitumen, cement, steel, aggregates, and so on) have increased 60 to 100 percent. Third, significant delays in project imple- mentation (which are not uncommon) are also associated with greater cost overruns, in part because they lead to greater exposure to other infla- tionary influences. The study concluded that cost overruns were the result of increased input costs against a growing demand for contracting in an environment of generally low competition for contracts. Hence, action is required to develop more competitive domestic markets for engineering contracting services. In view of the mounting upward pressure on road costs, it is relevant to ask whether any savings can be achieved by choosing alternative road technologies at the design stage. Key questions are whether the road surface type and condition are well aligned with the traffic volumes car- ried by each road, and whether the technologies used are the most cost- effective for delivering a particular type of surface. The RONET analysis shows a strong positive correlation between traffic levels and road surface type (that is, paved or unpaved), close to 0.7, although the correlation Roads: The Burden of Maintenance 63 between traffic levels and road condition is much weaker, ranging from 0.2 to 0.4. A minimum of 300 vehicles per day is widely accepted as the traffic threshold that makes paving of roads economically viable, and it is possi- ble to compare actual traffic levels against this benchmark. Paved roads with traffic volumes below the threshold have been potentially overengi- neered, while unpaved roads with high volumes are potentially under- engineered. On this criterion, there is some evidence of substantial overengi- neering in the main road networks, and much less of underengineering (figure 2.26). On average across countries, about 30 percent of the main networks appear to be overengineered and about 10 percent underengi- neered, suggesting a scope for significant cost savings by better aligning surface types with traffic volumes. Nevertheless, the variation across countries is huge. At one extreme, in Nigeria, almost 30 percent of the main road networks appear to be underengineered and only a minimal share are overengineered. At the other extreme, in Zambia, more than 60 percent of the main networks look to be overengineered. There are several possible explanations for overengineering. It may reflect a past expectation of high traffic growth that has not been realized, or a present expectation of high traffic growth in the near future. More commonly, however, it reflects political pressures (especially where cheap funding has been available) or a hope that maintenance performance, presently underfunded, will improve before periodic maintenance is required. On the rural network, the key traffic threshold is 30 vehicles per day, widely considered to be the minimum required to justify gravelling of roads. According to the minimum, 15 percent of the rural network length appears to be underengineered, meaning that a gravel surface is warranted (figure 2.27). At one extreme, in Burkina Faso, Ethiopia, Ghana, Lesotho, Mozambique, and Niger, 30 to 50 percent of the rural networks may be underengineered. At the other extreme, countries such as Chad, Rwanda, South Africa, Tanzania, and Uganda offer no evidence of underengineering. The results show no evidence of overengineering on the rural networks, implying no real scope for related cost savings. Further economies in road network costs could be made by adapting design standards to local conditions. Standards and warrants need contin- uous adjustment in light of materials availability and development. In turn, designs should take into account the local climate, natural materials available in the area, and traffic load and volume. In many cases, sealing gravel at traffic thresholds of less than 100 vehicles per day is economically 64 Co ng percentage of main road network Figure 2.26 o 70 80 90 100 0 10 20 30 40 50 60 M , Re a p Sw urit . az ius ila Ni nd ge So B ria ut en h in A Rw fric a a Ug nda an d To a g Ch o Et a hi d Se opi n a Ta eg nz al a Gh nia M ana a underengineered Cô G law te ab i Source: Calculation by A. Nogales based on data from Gwilliam and others 2009. d' on Iv Bu oire Ca ru Ga me nd m r i Bu bi oon rk a, T in h aF e a Ke so Extent of Over- and Underengineering on Main Road Networks ny N a Ce Le ige nt so r ra correctly engineered lA G tho fri Bo uin ca ts ea n w Re an pu a M b oz Su lic am da M b n ad iq ag ue as Co c ng m Zi M ar o, b al De ab i overengineered m we . Na Rep m . i Lib bia Za eria m Si E bia er rit ra re Le a on M Ang e au o rit la an ia 65 M percentage of rural road network Figure 2.27 M au 0 10 20 30 40 50 60 70 80 90 100 oz rit am ius b Le ique so t Gh ho an N a Bu Eth iger rk io in pi aF a Se aso ne Za gal m Ni bia g Na eri m a Cô K ib te en ia d' ya Iv M oir Ca ala e m wi er o M B on ad e Source: Calculation by A. Nogales based on data from Gwilliam and others 2009. Co ag nin ng asc o, ar Re To p. underengineered Ga go bo n Ce M nt An ali ra Bo go Extent of Over- and Underengineering on Rural Road Networks lA t fri B sw la ca u an n ru a Co Re nd ng pu i bl o, i De Ch c m ad .R Ga E ep. m rit correctly engineered bi re a, a T Gu he in e M Lib a au er rit ia Si Rw ani er a a ra nd Le a on Sw Sud e az an Ta ilan nz d U an Zi ga ia n So mba da ut b h we Af ric a 66 Africa's Transport Infrastructure justifiable even though the conventional standard is 200 vehicles per day. Sealed gravel roads have a black surface like any bitumen road. Typically, life-cycle cost savings would be on the order of 30 to 50 percent over 20 years compared with traditional surface treatments. The reduced cost of construction is achieved through reduced earthworks, reduced haulage distances for construction materials, reduced need for material processing, and reduced surfacing costs because of use of locally available materials. Pavement life is also increased because of reduced pavement deflection as pavement layers are compacted. Geometric standards also need review in light of the improvement of road materials. Prior to 2001, the de facto standard adopted in most Southern African Development Community (SADC) countries was the 1965 Policy on Geometric Design of Rural Highways--which did not cater specifically to low-volume roads--issued by the American Association of State Highway and Transportation Officials. More recently, the SADC has recognized that to minimize total transport costs, road improvements should be designed to meet the lowest practicable stan- dards (without unduly impairing safety requirements). Recognizing the shortcomings of using guidelines from developed countries, the United Kingdom's Transport Research Laboratory has pub- lished a series of Overseas Road Notes (TRL 2001) with more appropri- ate guidelines for developing countries. Guidelines have also been developed in Africa for use either nationally, for example, in South Africa (Transpotek 2001), or regionally (Pinard, Gourlay, and Greening 2003). The challenge is to apply existing designs and standards in a flexible man- ner to fit the parameters of the local environment and to do so safely and economically. To that end, the recent SADC guidelines offer advice on the implementation of low-volume sealed roads (SADC 2003). Labor-based methods have been an important part of the strategy to improve rural transport infrastructure in Africa over the past 35 years. These methods not only produce gravel roads of equal quality to those produced using equipment-based methods but also generate rural employment in a cost-effective manner. Nevertheless, these methods have not been applied on a large scale, often because of contractors' reluc- tance to adopt them (Stock 1996). First, contractors believe the cost of learning this new technology is high. Second, it has been argued that the cost of managing large labor forces makes labor-based methods more expensive than equipment-based methods. Unit-rate cost comparisons of labor-based and equipment-based methods, therefore, cannot predict firm behavior. Small firms appear more open to using labor-based methods Roads: The Burden of Maintenance 67 than large firms because they can supervise their sites more closely and increase worker productivity and control truancy more easily. Moreover, unlike large firms, small firms that wish to use equipment-based methods face high variable costs: they either own older, less-efficient equipment with high maintenance costs or must rent equipment at a high cost. Decentralization of responsibilities and improved financial manage- ment are essential for labor-based maintenance to work effectively. A review of experience gained under the Rural Travel and Transport Program in 1996 identified these as the two key reforms necessary to mainstream labor-based programs (Stock and de Veen 1996). Improved financial management is needed to ensure that funds flow adequately and laborers are paid on time, and decentralization is needed to streamline payment procedures and strengthen stakeholders' support of programs. These factors would need to be accompanied by strong government com- mitment, effective labor laws, appropriate design standards and training, and a suitable delivery mechanism. One way of assessing the burden of road maintenance at the country level is to look at the capital value of the road stock as a percentage of GDP (figure 2.28). In most countries, road networks are worth less than 30 percent of GDP. But some very poor countries (such as Malawi, Mozambique, Niger, and Zimbabwe), and countries with an exceptionally low population density (such as Namibia), have networks that are worth significantly more than that. It is in such countries that the fiscal burden of maintenance is likely to be particularly high. Identifying the Main Influences on Road Quality A key question is the extent to which road network quality is determined by economic and geographic fundamentals or can be influenced by pol- icy variables. GDP per capita has a significant statistical impact on the condition of main roads but, curiously, none whatsoever on the condition of rural networks. Overall, differences in the GDP alone explain 33 percent of the variation in road quality observed across countries. Nevertheless, both for main and rural roads there is a very wide range of network con- ditions across countries in the low-income bracket (with GDP per capita of less than $1,000 per year). Within the low-income class, the percentage of main roads in good condition ranges from 9 percent in Côte d'Ivoire to 74 percent in Burkina Faso. Similarly, the percentage of rural roads in good condition in the low-income countries ranges from 0 percent in Uganda to 63 percent in Burkina Faso. 68 percent Figure 2.28 0 20 40 60 80 100 Su d An an g Co G ola ng ab o, on R Ni ep. M ge a r Ca uri ia m tiu er s Se oon ne Cô Eth ga te iop l So d'I ia ut vo h ire Af ri Ke ca ny Be a Za nin m Road Asset Value as Percentage of GDP b Ch ia Rw ad Source: Calculation by A. Nogales based on data from Gwilliam and others 2009. Ta and nz a Ug ania Sw an a d Bo zila a ts nd w Co B an ng urk To a o, ina go De F m aso .R e M G p. ad ha M ag n oz a a am sc bi ar q Gu ue in e N a Le ige so r th o M Er ali itr Si M ea er al ra aw M Le i au on rit e Ce a nt Ga Na nia ra m mib lA bi ia fri a ca B , Th n uru e Re n pu di bl i Zi Lib c m er ba ia bw e Roads: The Burden of Maintenance 69 Geographic conditions also have a major impact on road conditions. In particular, countries with wetter and more mountainous terrain face sub- stantially higher costs of road construction and maintenance than do those with flat and arid terrain. A high rainfall level greatly accelerates the process of road deterioration, requiring frequent and more intensive maintenance interventions, and thus stretching the limited road sector budgets. A composite index is created that indicates the percentage of a country's national territory that is steep, moderately steep, or rolling and has rainfall in excess of 600 millimeters per year. The climate-terrain index shows a significant correlation with the quality of both main net- works and rural networks, though the correlation is stronger in the case of rural roads (figure 2.29). Vehicle overloading is without doubt one of the main influences on road quality. For example, as engineers estimate damage to road surfaces to be proportional to the fourth power of the axle weight, a road designed for a load of 9 tonnes per axle will incur 35 percent more damage per axle when the overload is only 10 percent. The aggregate cost of overloading for South Africa was estimated at $90 million a year in 1998. Many gov- ernments are trying to take action to reduce this cost. Kenya has attempted to ban heavy vehicles (through limiting the number of axles rather than the axle loads) and, in 2009, announced the intention of imposing heavy fines for overloading on owners rather than drivers. The South African Department of Transport drafted a National Overload Strategy in 2009, and there have been efforts within the regional eco- nomic communities to harmonize rules on overloading. Previous attempts to control overloading have not been successful, however. Control measures in Kenya have been challenged in the courts. While transit traffic can be controlled on entry to a country, domestic traffic is more difficult to control. Evasion by truckers has been extensive and systematic, aided by corrupt enforcement, sometimes at a high level. For example, Trans-African Concessions, which runs the motorway between Maputo and Witbank in South Africa, has complained that the Mozambican police and the National Road Administration are not doing enough to stop overloading of trucks on this road. Given the practical difficulties of adjusting trucks to the roads, it may be sensible to consider the converse policy of adjusting the roads to the trucks. The economies of scale of heavier trucks are compelling, and a carrier has strong incentives to load his vehicle to the maximum. Studies in the mid-1980s showed that the savings in operating costs when trucks are allowed to carry 12­15 tonnes per single axle, rather than the usual 70 Africa's Transport Infrastructure Figure 2.29 Relationship between Road Networks in Poor Condition and the Climate-Terrain Index a. Main roads 100 climate-terrain challenge index (%) 90 80 70 60 50 2 R = 0.19 40 30 20 10 0 0 20 40 60 80 100 percentage of main roads in poor condition b. Rural roads 100 climate-terrain challenge index (%) 90 80 70 60 50 2 R = 0.24 40 30 20 10 0 0 20 40 60 80 100 percentage of rural roads in poor condition Source: Gwilliam and others 2009. 8­10 tonnes, far outweigh the extra cost of constructing or repaving roads to bear the heavier load (World Bank 1988). This finding implies that on all but lightly used road networks, stronger pavement is economically jus- tified. Moreover, as the worst offenders are usually dump trucks hauling the densest of cargo--crushed stone, sand, gravel, cement--specific, tar- geted regulations requiring multiaxle vehicles for these businesses might be a more enforceable policy. Roads: The Burden of Maintenance 71 The problem is that upgrading a whole network designed for relatively low-axle loads could be extremely expensive. In summary, though vehicle overloading adversely impacts road conditions, this does not necessarily imply that restricting vehicle size is the appropriate policy response. Requiring large vehicles to have multiple axles is one alternative. Recognizing that axle-load limits are difficult to enforce, a policy emphasis on strengthening roads to achieve the operating cost savings associated with very heavy vehicles is another. Detailed analysis of this long-term strategic decision should be a high priority. Institutional arrangements also matter. Countries with both a road fund and a road agency have 20 percent more of their main and rural road networks in good or fair condition than countries without these two elements. The quality of the road fund institutions, as measured by a road fund quality index devised for this study, also has a substan- tial and significant effect on the percentage of the main road networks in good condition but not on the quality of the rural road networks (figure 2.30). In countries with high fuel levies, an additional 10 percent of the main road networks and an additional 5 percent of the rural road networks are in good or fair condition. (But there is no clear ranking of countries with low fuel levies versus no fuel levies at all.) As might be expected, the level of maintenance expenditures shows strong correlation with the quality of the main networks but not with that of the rural networks (figure 2.31). Policy choices on road institutions and funding levels thus have a mate- rial impact on the quality of the main road networks. Countries with both road funds and road agencies, as well as those with high fuel levies and relatively high maintenance expenditures, seem to reap the benefits and have a higher proportion of their main road networks in good or fair con- dition. But these variables have a much weaker impact on the quality of the rural road networks. This situation may reflect deficiencies in the accuracy of data on spending and road quality for the rural networks, or it may reflect the fact that rural network management is driven by insti- tutions and resource allocations at the local level, and thus does not ade- quately reflect national policy. Freight Transport: Too Expensive Freight transport services are very important to the African economies, many of which are dependent on exports of relatively low value-for- weight goods to world markets. Unfortunately, empirical studies carried 72 Africa's Transport Infrastructure Figure 2.30 Relationship between Road Networks in Good Condition and Their Score on the Road Fund Quality Index a. Main roads 100 2 R = 0.41 score on road fund index (%) 80 60 40 20 0 0 20 40 60 80 100 percentage of main roads in good condition b. Rural roads 100 2 R = 0.06 score on road fund index (%) 80 60 40 20 0 0 20 40 60 80 100 percentage of rural roads in good condition Source: Gwilliam and others 2009. out since the mid-1990s have consistently demonstrated that transport prices in Africa are higher than in other regions. Rizet and Hine (1993) estimated that prices of road freight transport in Cameroon, Côte d'Ivoire, and Mali were six times those in Pakistan. A later study (Rizet and Gwet 1998) demonstrated that for distances up to 300 km, unit costs of road transport were 40 to 100 percent higher in Africa than in Southeast Asia. Transport charges for landlocked African countries have been shown to range from 15 to 20 percent of import costs, a rate that is Roads: The Burden of Maintenance 73 Figure 2.31 Relationship between Road Networks in Good Condition and Maintenance Expenditures maintenance spending per kilometer (US$) a. Main roads 7,000 6,000 5,000 2 R = 0.30 4,000 3,000 2,000 1,000 0 0 20 40 60 80 100 percentage of main roads in good condition maintenance spending per kilometer (US$) b. Rural roads 7,000 6,000 5,000 4,000 3,000 2 R = 0.00 2,000 1,000 0 0 20 40 60 80 100 percentage of rural network in good condition Source: Gwilliam and others 2009. three or four times as high as that typically found in developed countries (MacKellar, Wörz, and Wörgötter 2000). There is also substantial variation within Africa. While variable trans- port costs per vehicle-km generally fall in the range of $1.23 to $1.83, fixed costs, transport quality, and transport journey speeds vary. Transport time for long journeys is itself a good indicator of quality of service. As table 2.10 shows, transport quality in South Africa, where larger trucks are used, is higher than in Central or West Africa, but, on average, prices 74 Africa's Transport Infrastructure Table 2.10 Performance of the International Gateway Corridors Price per Price per Distance Transit tonne tonne-km Region Gateway Destination (km) time (days) (US$) (US$) East Mombasa Kampala 1,100 5­6 90 .081 Mombasa Kigali 1,700 8­10 100­110 .059­.065 West Lomé Ouagadougou 1,050 6­8 60­70 .057­.067 Cotonou Niamey 1,000 6­8 65­95 .065­.095 Central Douala Ndjamena 1,850 12­15 200­210 .108­.113 Douala Bangui 1,450 8­10 200­210 .138­.145 South Durban Lusaka 2,300 8­9 90­130 .039­.057 Durban Ndola 2,700 9­10 130­170 .048­.063 Source: Teravaninthorn and Raballand 2008. are lower. A recent study (Teravaninthorn and Raballand 2008) showed that, on average, transport prices in Central Africa are between two and three times as high as those in southern Africa. High freight charges in Africa were initially attributed to the effects of infrastructure constraints on vehicle operating costs (Limão and Venables 2001). Certainly the structure of trucking costs in Africa differs consider- ably from that in most regions of the world. Compared to European oper- ators, African truckers tend to have low fixed costs (resulting from low salaries and the use of cheap, old trucks) and high variable costs (mainly attributable to the high fuel consumption of these old and poorly main- tained trucks). Poor road conditions reduce the life of trucks and tires, increase vehicle maintenance costs, and increase fuel consumption. In fact, fuel and lubricants account for between 40 and 70 percent of the total variable costs. In general, variable costs account for an unusually large proportion of total costs (over 70 percent in Central and West Africa); as a consequence, the incentive to make intensive use of the vehicles is weakened. The age of the truck fleet and the low utilization of vehicles seem to be even more critical than the unit cost of inputs. Annual truck mileage is lower in many Central, East, and West African countries than in developed countries and many other developing countries. For exam- ple, average truck mileage is less than 70,000 km per year in Cameroon, Ethiopia, Malawi, and Niger, compared with about 110,000 km in Pakistan and South Africa. More recent studies have tended to emphasize the institutional and regulatory influences on freight charges. Teravaninthorn and Raballand (2008) have shown that the range of transport costs is less than the Roads: The Burden of Maintenance 75 range in prices, with variable operating costs varying by a factor of 0.5 but prices varying by as much as a factor of 4. For example, the average transport price per tonne-km ranges from 4 to 5 cents in southern Africa, from 6 to 8 cents in West and East African corridors, and from 10 to 25 cents in the Central African corridor. This finding suggests that there might be significant degrees of monopoly in the higher price markets. At first glance, that explanation would appear unlikely, as the freight transport industry is generally fragmented in West and Central Africa, with virtually no large trucking companies in the business. But the small operators are typically tightly regulated by freight bureaus, shippers' councils, and trade unions. And the devices used to ensure an equitable distribution of income among operators--most notably "tour de role" dis- patching--increase costs.4 This increase is accentuated by the degree to which restricted competition between haulers allows monopoly profits to be taken. East Africa has a more competitive and mature market. There are about 20 large companies with more than 100 trucks each on the main East African corridors, and the largest Kenyan company has a fleet of over 600 vehicles. The large companies account for about 20 percent of the total market--a figure comparable to that found in Europe and North America. Southern Africa also has a more mature structure, particularly in the regulatory and logistical arenas. Differences in market power thus account at least in part for the high prices in some regions. That is not all, however. A peculiarity of African transport is that, con- trary to experiences in the rest of the world, the price per tonne-km for long-distance freight destined for international markets is higher than that for domestic traffic within a country. Procedures used by customs and border-crossing officials contribute to the low annual vehicle usage fig- ures: there seems to be a strong positive correlation between transport prices and the number of border crossings. This suggests that there are serious deficiencies in the regulatory regime relating to transit traffic, sub- stantiated to some extent by the existence of high profit margins in inter- national movements. These high profits are achieved despite low annual utilization of vehicles and many nontariff barriers in Central and West Africa. The most plausible reason for this peculiarity lies in the role of the official and nonofficial regulation of the sector. Government-imposed procedures also contribute to high freight charges. International traffic is strictly governed in both West and Central Africa by bilateral transit agreements, implemented by national freight 76 Africa's Transport Infrastructure bureaus. Quotas are set on the proportion of each trade that can be carried by the party countries, and cabotage is banned. The freight bureaus are able to use their formal powers to manage the issuance of cargo- and tran- sit-related documents to act as monopolist freight allocation bureaus and are instrumental in maintaining high freight rates in collaboration with the truckers' unions. This finding is supported by customer research. The perception of international freight forwarders, expressed in the World Bank's Logistics Performance Index, place the four African subregions below all other regions of the world, with southern Africa the best of the four and West Africa the worst (World Bank 2010). Operations to and from South Africa are governed by bilateral agree- ments, which provide for a sharing of information on traffic development and define the types of permits that can be issued. This system restricts the carriage of bilateral trade to operators from the two countries con- cerned and prohibits cabotage. But it does not establish quotas, and it allows rates to be determined by the market to enable direct contracting between shippers and transporters and to give incentives to efficient operators. The southern African international transport market is a good model for the rest of the continent because it combines liberalization of entry with enforcement of quality and with load control rules applicable to all operators. Currently, around 70 percent of the main trade corridors are in good condition, and donors are increasingly channeling resources to infra- structure improvements along these strategic routes. But there is also recognition that it will take more than good infrastructure to make these corridors function effectively. Neighboring countries have increas- ingly organized themselves into corridor associations to address the nonphysical barriers to transit, with a particular focus on cutting lengthy delays (between 10 and 30 hours at border crossings and ports) by creating one-stop integrated frontier posts and improving ports and customs administration. The southern African corridor performs significantly better than those in Central and West Africa, approaching developing-country norms in terms of freight tariffs; but even here, the duration of transit leaves much to be desired. Notwithstanding the emphasis on trade facilitation, the AICD analysis indicates that the high cost and low quality of road freight service in Central and West Africa is primarily attributable to a highly reg- ulated and cartelized trucking industry, making liberalization the number one priority to improve road transport in that region. Roads: The Burden of Maintenance 77 The Way Forward Africa's road network, though physically sparse, is relatively large com- pared to the size of its population, and even larger when seen in the con- text of national income. Countries, on average, spend around 2 percent of their GDP on roads. Within this envelope, there is a significant bias toward capital expenditure. This bias is further exacerbated when one considers that countries are typically able to execute only around 60 per- cent of budgeted capital spending. As a result, countries are budgeting, on average, only 30 percent of road expenditure to maintenance, versus a norm of more than 50 percent in more mature road systems. Nevertheless, even with this degree of capital bias, only about half of the countries surveyed have capital expenditures large enough to clear cur- rent network rehabilitation backlogs within a reasonable time period. At the same time, fewer than half of the countries are allocating enough resources to cover routine and periodic maintenance requirements. As a result, a significant number of countries are in a vicious cycle of low maintenance budgets leading to network deterioration leading to an esca- lating rehabilitation backlog--a backlog that they lack adequate capital resources to clear. Recent escalations of unit costs for road maintenance and reconstruction are likely to further dilute the adequacy of road budget allocations. The policy response to this situation has been the widespread adoption of second-generation road funds, though not all have been well designed or well implemented. In many countries, the fuel levy is too low, and in some, collection of the levy has posed a serious problem. Nevertheless, countries with road funds--in particular those that also set fuel levies at a reasonably high level--have systematically better road financing, exhibit a lower degree of capital bias, and are much closer to covering road main- tenance requirements. While income and geographical factors have a sig- nificant impact on the condition of networks, quasi-independent road funds and road agencies are also highly beneficial. Lack of funding and institutional capacity shows up most strongly in the condition of the unpaved and lower tiers of the network. There is thus a need to spend as cost-effectively as possible, in particular by exploring the potential for cost savings through the adoption of more appropriate technological standards. Even within the current technology, there is evi- dence of substantial overengineering of the main road networks relative to traffic volumes. The rural networks, on the other hand, tend to be somewhat underengineered. Road transport operations, though private, 78 Africa's Transport Infrastructure are costly and relatively inefficient, with a lack of competition accentu- ated by poor administrative procedures for allocation of traffic, and inef- ficiency and corruption at ports and land borders. The priorities for the future all stem from this analysis. Priority 1. Consolidate road-funding arrangements. Countries with (well-financed) road funds have been shown to do sig- nificantly better at capturing resources for maintenance than those without them. But the quality of the administrative arrangements makes some road funds more successful than others. For consolidation of the gains already made in the region, the following suggestions should be heeded: · Countries without second-generation road funds should establish them immediately. · All road funds should be founded in law rather than by administrative decree, should provide for direct transfer of levy revenues to the fund, and should have majority user participation in managing the road fund board, with published auditing. · Governments should require the road fund board to demonstrate what level of fuel levy or other revenue source is necessary to prevent deterioration of the network, and what is necessary to overcome back- logs in maintenance over a reasonable period. · Road boards should be required to develop transparent formulas or procedures to govern the allocation of road fund revenues to differing road categories. Priority 2. Commercialize maintenance implementation arrangements. The development of commercially structured road authorities, independ- ent of direct ministerial control, has improved performance and facili- tated the introduction of new procedures in several countries. It is therefore recommended as a parallel approach to road maintenance and includes the following aspects: · Establishment of quasi-autonomous road authorities with user repre- sentation on boards · Introduction of performance-based road maintenance contracts · Development of information and training programs for road mainte- nance contractors. Roads: The Burden of Maintenance 79 Priority 3. Make a concerted effort to improve road safety. Africa has the worst road accident record in the world. Programs to reduce accidents have succeeded elsewhere. Recommendations include · Establishing a high-level national safety council · Conducting safety audits on all new road and road improvement designs · Developing a comprehensive national road safety program. Priority 4. Liberalize the road haulage sector. Cartelization of operations and failures in market regulation (including the enforcement of a "tour de role" dispatching for import traffic) limit the competitiveness and hinder the efficiency of the trucking industry. To overcome these problems, countries should consider the following recommendations: · Legislation should be introduced restricting entry to road haulage markets on qualitative conditions only (including operator, vehicle, and driver licensing). · Road haulage associations should be excluded from the setting of prices or the allocation of traffic among members. · Liberal approaches should be adopted toward foreign haulers involved in cabotage markets. Notes 1. The main source document for this chapter is Gwilliam and others (2009). The Road Network Evaluation Tool (RONET) analysis was done by Alberto Nogales. Source materials from the Sub-Saharan Africa Transport Policy Program documentation include Stock (1996) and Stock and de Veen (1996). Heggie and Vickers (1998) and Malmberg Calvo (1998) provided important information on road management issues. 2. In chapters 7 and 8, the low-income, aid-dependent countries are further sub- divided into "fragile" and "nonfragile." For more on this classification see World Bank (2007b). 3. Note that the totals vary from 100 percent because of rounding. 4. "Tour de role" dispatching involves vehicles queuing at the dispatch point, with work allocated to vehicles strictly in accordance with their position in the queue. 80 Africa's Transport Infrastructure References African Development Bank. 2003. "Review of the Implementation Status of the Trans-African Highways and the Missing Links." SWECO International AB and Nordic Consulting Group AB, Stockholm. Benmaamar, M. 2006. "Financing of Road Maintenance in Sub-Saharan Africa: Reforms and Progress towards Second Generation Road Funds." Sub-Saharan Africa Transport Policy Program Discussion Paper 6, World Bank, Washington, DC. Briceño-Garmendia, C., K. Smits, and V. Foster. 2009. "Financing Public Infrastructure in Sub-Saharan Africa: Patterns, Issues, and Options." Africa Infrastructure Country Diagnostic Background Paper 15, World Bank, Washington, DC. Carruthers, R., R. R. Krishnamani, and S. Murray. 2009. "Improving Connectivity: Investing in Transport Infrastructure in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 7, World Bank, Washington, DC. Gwilliam, K., and A. Kumar. 2003. "How Effective Are Second-Generation Road Funds? A Preliminary Appraisal." World Bank Research Observer 18 (1): 113­28. Gwilliam, K., V. Foster, R. Archondo-Callao, C. Briceño-Garmendia, A. Nogales, and K. Sethi. 2009. "The Burden of Maintenance: Roads in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 14, World Bank, Washington, DC. Heggie, I. G., and P. Vickers. 1998. "Commercial Management and Financing of Roads." World Bank Technical Paper 409, World Bank, Washington, DC. Limão, N., and A. J. Venables. 2001. "Infrastructure, Geographical Disadvantage, and Transport Costs." World Bank Economic Review 15 (3): 451­79. MacKellar, L., J. Wörz, and A. Wörgötter. 2000. "Economic Development Problems of Landlocked Countries." Transition Economics Series 14, Institute of Advanced Studies, Vienna. Malmberg Calvo, C. 1998. "Options for Managing and Financing Rural Transport Infrastructure." World Bank Technical Paper 411, World Bank, Washington, DC. Pinard, M. I. 2009. "Review of Progress on the Commercialization of Road Agencies in Sub-Saharan Africa." Sub-Saharan Africa Transport Policy Program, World Bank, Washington, DC. Pinard, M. I., C. S. Gourlay, and P. A. K. Greening. 2003. "Rethinking Traditional Approaches to Low-Volume Road Provision in Developing Countries." Transportation Research Board Paper LVR8-1153, Transportation Research Board of the National Academies, Washington, DC. Rizet, C., and R. Gwet. 1998. "An International Comparison of Road Haulage Prices in Africa, South-East Asia and Latin America." Recherche Transports Sécurité 60: 69­85. Roads: The Burden of Maintenance 81 Rizet, C., and J. Hine. 1993. "A Comparison of Costs and Productivity of Road Freight Transport in Africa and Pakistan." Transport Reviews 13 (2): 151­65. Robinson, R., and D. Stiedl. 2001. "Decentralisation of Road Administration: Case Studies in Africa and Asia." Public Administration and Development 21 (1): 53­64. SADC (Southern African Development Community). 2003. "Guideline on Low- Volume Sealed Roads." SADC, Gabarone, Botswana. SSATP (Sub-Saharan Africa Transport Policy Program). 2004. "RMI Matrix for August 2004." World Bank, Washington, DC. ------. 2007. "RMI Matrix for September 2007." World Bank, Washington, DC. Stankevich, N., N. Quereshi, and C. Queiroz. (2005). "Performance-Based Contracting for Preservation and Improvement of Road Assets." Transport Note TN-27, World Bank, Washington, DC. Stock, E. A. 1996. "Problems Facing Labor-based Road Programs and What to Do About Them--Evidence from Ghana." Sub-Saharan Africa Transport Policy Program Working Paper 24, World Bank, Washington, DC. Stock, E. A., and J. de Veen. 1996. "Expanding Labor-Based Methods for Road Works in Africa." Sub-Saharan Africa Transport Policy Program Working Paper 22, World Bank, Washington, DC. Teravaninthorn, S., and G. Raballand. 2008. Transport Prices and Costs in Africa: A Review of the Main International Corridors. Washington, DC: World Bank. Transpotek. 2001. G2 Geometric Design Manual. Pretoria: Council of Scientific and Industrial Research. TRL (Transport Research Laboratory). 2001. "Management of Rural Road Networks." Overseas Road Note 20, TRL, Crowthorne, Berkshire, U.K. World Bank. 1988. Road Deterioration in Developing Countries: Causes and Remedies. Washington, DC: World Bank. ------. 2007a. "Rural Accessibility Index: A Key Development Indicator." Transport Paper 10, World Bank, Washington, DC. ------. 2007b. "Index for Countries Eligible for World Bank Loans and IDA Credits." Internal Memo 10, World Bank, Washington, DC. ------. 2008a. Privatization Database. World Bank, Washington, DC. http://rru.worldbank.org/Privatization. ------. 2008b. ROCKS (Road Costs Knowledge System) (database). World Bank, Washington, DC. http://go.worldbank.org/ZF1I4CJNX0. ------. 2009. RONET (Road Network Evaluation Tool) (software). World Bank, Washington, DC. http://go.worldbank.org/A2QQYZNFM0. ------. 2010. Logistics Performance Index (software). World Bank, Washington, DC. http://go.worldbank.org/88X6PU5GV0. CHAPTER 3 Railways: Not Pulling Their Weight Railways transformed the face of Africa in the late 19th and early 20th centuries, creating strategic corridors that opened the interior for the exploitation of mineral and other resources. But most lines remain iso- lated, with little network interconnection. Built to modest technical stan- dards, railways were left unprepared to compete for time-sensitive traffic (including passengers) as road systems developed. Revenues have been generally insufficient to finance the modernization of track and rolling stock. Conservative management under state ownership has not helped, and facilities have suffered disproportionately in postindependence civil wars. While concessioning to the private sector promises to improve oper- ational efficiency, the railways still face serious financial problems. New forms of partnership between states and the private sector are needed if the rail sector is to be revitalized. Africa's Rail History: Opening Up the Continent The first railways south of the Sahara were built in South Africa in the 1860s and 1870s, with lines heading inland from the ports at Cape Town and Durban.1 While railways continued to develop in Cape Province, Natal, and Transvaal, it was not until the turn of the 20th century that 83 84 Africa's Transport Infrastructure large-scale railway development began in other parts of the continent. In most cases, the development consisted of isolated lines heading inland from a port to a trading center or mine, with a few branch lines added later. Many lines were state owned, but some were managed as conces- sions or constructed by mining companies. In the past 80 years, few lines have been constructed outside South Africa and its immediate neighbors. The most significant is the Tazara line, built by the Chinese during the 1970s, which links Tanzania and Zambia. Other major projects include the Trans-Gabonais (opened in 1987 principally to transport minerals), the extension of the Cameroon network from Yaoundé to Ngaoundere, and the northeastern extension of the Nigerian network from Kuru to Maiduguri. Although there have been grand network plans for over a century, most railways in Africa consist of disconnected lines, either within coun- tries or linking ports to their regional hinterlands. The only true interna- tional networks are those centered on South Africa and stretching north to Zimbabwe, Zambia, and the Democratic Republic of Congo, and, to a lesser extent, the old East African railways network in Kenya, Tanzania, and Uganda. This pattern of railway development reflects the historically limited amount of intercountry trade in Africa. Even today, trade volumes between adjacent countries are remarkably small. African railways are therefore closely linked to ports (in fact, much of Africa once had inte- grated port and railway organizations). Where railways traverse more than one country, freight rarely originates or terminates in the intermediate country or countries--with the notable exceptions of traffic between Kenya and South Africa and their neighbors (map 3.1). Most railways in the region were reasonably successful until the 1960s. But as the road system developed in Africa, new and larger trucks increas- ingly captured the higher-value general freight. Rail traffic became lim- ited primarily to bulk mineral and agricultural freight and semibulk freight such as fuel. The resulting decrease in revenues delayed the main- tenance and replacement of deteriorating track and rolling stock. Therefore, even when railways tried to reclaim higher-value traffic (such as containers), their low quality of service prevented them from taking a significant market share from road competition. Other factors contributed to the decline of railways in Africa. For example, governments required railways to operate unprofitable passen- ger services without compensation. This practice not only drew cash away from infrastructure improvements, but also tied up locomotives that could have been used for revenue-generating freight services. The many Railways: Not Pulling Their Weight 85 Map 3.1 The African Rail Network in 2009 PORT SUDAN DAKAR DJIBOUTI CONAKRY ABIDJAN TEMA PORT DOUALA LOME HARCOURT LAGOS MOMBASA POINTE NOIRE DAR ES SALAAM LUANDA BEIRA TOAMASINA Major port WALVIS BAY Railway MAPUTO DURBAN EAST LONDON CAPE TOWN PORT ELIZABETH Source: Bullock 2009. wars and civil disturbances in Africa over the past 50 years have also hin- dered railway development--either directly, through the destruction of facilities (Angola, Eritrea, Ethiopia, and Mozambique), or indirectly, by cutting inland railways off from their ports (Burkina Faso and Malawi). A Sparse and Disconnected Network The rail system of Africa comprises various lines and small networks that, combined, offer low-density coverage and little interconnection between regions. At the end of 2008, there were 52 railways operating in 33 coun- tries in Africa (appendix 3a). Most of these used one of two rail gauges2: Cape gauge (1,067 millimeters, or 3 feet 6 inches) or meter gauge. 86 Africa's Transport Infrastructure The total African network size is around 70,000 kilometers (km), of which about 55,000 km is currently being used. The network is single track except for sections of the Spoornet network (recently renamed Transnet Freight Rail). Very little is electrified outside of South Africa (where 42 percent of the network--nearly 9,000 km--is electrified); the only other electrified sections are 858 km in the Democratic Republic of Congo and 313 km in Zimbabwe (not currently in use). The spatial density of the rail network in Africa is low. The highest den- sity is in South Africa (16 route-km per 1,000 square km [km2]), compared to a range of 1 to 6 route-km per 1,000 km2 in most other African coun- tries. Australia, Canada, China, and Russia, all of which have large undevel- oped and sparsely populated areas, have densities between 5 and 7, while densities for most European countries range from 20 to 100. Thirteen coun- tries in Africa do not have operating railways. Network density with respect to population, measured in route-km per million people, is highest in Gabon (520), Botswana (494), and South Africa (460). Most other African countries have densities ranging from 30 to 150 route-km per million peo- ple (see appendix 3a). In comparison, densities in European countries range from 200 to 1,000, and reach over 1,500 in Australia and Canada. The main interconnected network in southern and Central Africa-- which extends as far north as the Democratic Republic of Congo and southern Tanzania--uses Cape gauge. The same is true for railways in the ex-British possessions of Ghana, Nigeria, and Sudan. Meter gauge is used in all other former French possessions; in the disconnected Ethiopian line; and in the East African network that links Kenya, Uganda, and northern Tanzania. There are also a number of isolated standard gauge lines: those in Guinea and Mauritania are privately operated mineral lines; the stan- dard gauge line in Gabon, although developed primarily for mineral traf- fic, is a public railway, which also carries general traffic and offers passenger services; and Eritrea has the only narrow gauge line in Africa. Despite the variety of rail gauges, interoperability of railways is not a major problem. In only three places--two in Tanzania and one in Guinea--are there two different gauges in the same location. The Cape gauge network based in South Africa connects 11 countries, and the East African network connects 3. Two international meter gauge networks in West Africa connect landlocked Francophone countries to the coast: Ouagadougou-Abidjan (Sitarail), which links Burkina Faso to Côte d'Ivoire, and Bamako-Dakar (Transrail), which links Mali to Senegal. Another meter gauge network in East Africa links Ethiopia to Djibouti. Some other net- works that do not cross international borders provide railheads from which Railways: Not Pulling Their Weight 87 traffic can continue by road. For example, in Benin the Organisation Commune Benin-Niger des Chemins de Fer et des Transports (OCBN) provides a link from Cotonou to Niamey through a railhead at Parakou; Camrail provides railheads for traffic between the port of Douala in Cameroon and the Central African Republic and Chad; and in East Africa, Tanzania Railways Corporation (TRC) carries traffic for Burundi and east- ern Democratic Republic of Congo, and Kenya Railways Corporation (KRC) for Rwanda. Over 50 companies operate in Africa. Many are small, although a single company, South Africa's Transnet Freight Rail, has about 40 percent of the operating network and carries 70 percent of the traffic. South Africa also dominates the rail passenger business. Mine-connected rail lines in both West and southern Africa constitute only 4 percent of the network but carry over half the freight (as measured by net tonne-km [ntkm]), most of which is carried on the Transnet Freight Rail coal and ore export lines. There are other mineral lines in Gabon, Guinea, Mauritania, and Nigeria, some of which also carry general traffic to and from mines. Traffic Volume: Unprofitably Low Traffic volumes on the region's railways are generally low by world stan- dards. South Africa's rail system averages around 5 million traffic units (TUs) per route-km overall3--this figure is only 2.4 million when special- ized coal and ore lines are excluded. The network with the next highest average is that of Gabon, with 2.7 million TUs per route-km. Cameroon's Camrail (1.1 million) is the only other railway with an aver- age density of over 1 million TUs per route-km (see appendix 3b). Many railways average under 300,000. Even in South Africa, only 50 percent of the networks carry more than 2 million net tonnes per year. With such low traffic volumes, many networks in Africa struggle to maintain and renew their infrastructure. Infrastructure Condition: Impeding Rail's Competitive Potential Most networks in Africa, outside of South Africa, still operate at the stan- dards to which they were constructed. They are small-scale, undercapital- ized networks designed for relatively low axle loads and low speeds, ill-suited to modern requirements. The rail track itself is often too light for even the moderate axle loads currently being operated. When the Dakar-Senegal railway was conces- sioned, the average age of track was reported as 37 years in Senegal and 51 years in Mali. Most track is even older. In addition, the strength of rail 88 Africa's Transport Infrastructure manufactured 60 or 70 years ago is often well below current standards, leading to fatigue failures and rail fractures. Additional difficulties arise in countries emerging from conflict. For example, in Angola and Mozambique, most infrastructure was destroyed in conflicts; mines had to be removed before rail lines could even be rehabilitated. Control systems are also archaic. Many networks still rely on mechani- cal signals and train orders. While on most lines these systems are adequate from a capacity standpoint, human error often causes significant safety problems. Unfortunately, power signaling, where installed, often does not operate because of short circuits, lack of electrical power, and dilapidated cable networks. Rail telephone exchanges are similarly obsolete, having limited capacity and requiring spare parts that are virtually impossible to find. In addition, many structures, such as bridges and viaducts, are now over 100 years old. (See appendix 3a for more on railway assets.) Because of chronic undermaintenance, many sections of the aging track have deteriorated, almost beyond repair. In most networks, considerable sections of track require repair or replacement. Major sections are inopera- ble in several countries, including Benin (23 percent), Angola (69 percent), and Uganda (91 percent). These networks will require rehabilitation before operations can recommence. In other countries, much of the network is not used on a regular basis (up to 60 percent in Ghana). Where services are operated, poor track conditions restrict train speeds on long sections, which reduces railway competitiveness and rolling-stock productivity. The cost of repairs is beyond the financial capacity of most railways based on current traffic volumes. Conservative estimates put repair costs at $200,000 per kilometer in the most straightforward cases, and probably closer to $350,000 per kilometer on average. Funding repairs at these costs would absorb all operating surpluses for many years, by which time another backlog will have appeared. Rehabilitation is unlikely to be eco- nomically justified for many sections unless they show good prospects for bulk traffic or have no road competition. Lines carrying less than 1 million net tonnes per year are unlikely to warrant major rehabilitation, and lines carrying under 250,000 net tonnes per year probably cannot support any- thing more than routine maintenance. Network Expansion Proposals: Often Lacking Economic Focus There have been many proposals, some dating back a century, to create new routes for landlocked countries and to integrate the isolated networks. The most ambitious proposal came in 1976, when the African Railways Union prepared a master plan for a Pan-African rail network, which Railways: Not Pulling Their Weight 89 included 18 projects requiring 26,000 km of new construction. The plan was approved by the Organization of African Unity in 1979. In 2001, the African Railways Union published a revised master plan containing a sub- set of 10 corridors and, in 2005, further simplified the plan into three major transcontinental routes: · Libya­Niger­Chad­Central African Republic­the Republic of Congo­ the Democratic Republic of Congo­Angola­Namibia (6,500 km) · Senegal­Mali­Chad­Djibouti (7,800 km) · Kenya­Tanzania­Uganda­Rwanda­Burundi­the Democratic Republic of Congo, with possible extensions to Ethiopia and Sudan (5,600 km) There have also been proposals for individual lines: a link from Isaka in Tanzania to Rwanda, with complementary links from Rwanda and Burundi to the Ugandan and Tanzanian network; a link through Kenya (or possibly Uganda) to southern Sudan; an extension of the Lilongwe line in Malawi; and a route from Walvis Bay in Namibia to Zambia and Angola. Company mineral lines have been proposed in Gabon, Mozambique, Namibia, and Sierra Leone. But few, if any, of the proposed links have moved beyond the drawing board. In practice, some recent proposals have a clearer economic focus. For example, China is increasingly interested in oil and precious minerals in Angola, the Democratic Republic of Congo, and Zambia, and is investing in a new railway line to assist in the extraction and transportation of these resources to China. The core of the plan is the 1,860 km long Tanzania- Zambia Railway, which links the Indian Ocean port city of Dar es Salaam in Tanzania to Kapiri Mposhi, Zambia. This was built by the Chinese gov- ernment in the 1970s, primarily to free Zambia from a politically based trade blockade by what was then Southern Rhodesia. But its traffic has diminished and it now has only about 20 percent of its original locomo- tive capacity. In 2001, China pledged to finance development of this rail line to create a railway crossing the continent from coast to coast. The Chinese plan would extend the line through the southern part of the Democratic Republic of Congo and link it with the current Chinese development in Angola. Investment and Maintenance With such low traffic volumes, African railways face a continuing problem in financing either new investment or maintenance of the existing system. 90 Africa's Transport Infrastructure Infrastructure: Difficult to Finance on Low-Volume Lines The capital cost of new rail infrastructure is high. The construction of a single-track, nonelectrified railway costs at least $1.5 million per kilome- ter in relatively flat terrain, and around $5 million in more rugged coun- try requiring more extensive earthworks. The reconstruction of an existing line for which the right of way and earthworks already exist typically costs about $350,000 per kilometer using new materials; if secondhand materials such as cascaded rail can be used, the cost is lower--around $200,000 per kilometer. For lines that are to be upgraded, bridges may require strengthening to handle higher axle loads. Additional earthworks may also be required if alignments are to be improved. Hence, as a rule of thumb, the cost of upgrading can easily be twice the cost of simple track renewal. These costs exclude signaling, for which relatively cheap options are now available for the typical low-volume network. Periodic replacement also imposes significant costs. Even if recon- structed infrastructure has a useful life of 40 to 50 years, the annual cost of maintenance is $5,000­$10,000 per kilometer, excluding any return on investment. If a low 5 percent return on investment is included, the annual cost increases to $20,000­$40,000 per kilometer. This means that a line that carries 1 million net tonnes per year would need to earn 0.5­0.8 cents per net tonne-km (c/ntkm) to fund periodic rehabilitation, while traffic with a density of 250,000 tonnes per year would need to earn 2­3 c/ntkm. If return on investment is included in the cost, the required returns quadruple to 2­3 c/ntkm for lines carrying 1 million net tonnes per year. Yields on most freight railways in Africa are around 4­5 c/ntkm and oper- ating costs are 3­4 c/ntkm, leaving at most 1 c/ntkm for rehabilitation. Therefore, unless the investment creates a significant increase in traffic, full rehabilitation is commercially viable only for those lines with a density of 2 to 3 million net tonnes or more. Investment Needs A full analysis of the investment needs of railways in Africa would require detailed data on the conditions of infrastructure and rolling stock and traffic volumes for each railway. In the absence of such data, rough esti- mates have been made using aggregate statistics and broad assumptions. The railway network in Africa north of South Africa and south of the Sahara consists of about 44,000 km of track, of which about 34,000 km is currently operational. Nearly all the lines are low volume and would thus justify only partial rehabilitation, possibly using cascaded materials. Railways: Not Pulling Their Weight 91 Even assuming a relatively low unit cost--around $200,000 per km-- probably no more than 15,000­20,000 km of the network can support this level of expenditure. Over a 40-year interval, the cost of infrastruc- ture rehabilitation would therefore average around $100 million per year. The cost of replacing rolling stock can be estimated in a similar man- ner. The railway network north of South Africa carries around 15 billion ntkm per year (excluding the mineral lines) and about 4 billion passenger- km (pkm). On average, 500 wagons, 20 passenger carriages, and about 20 locomotives will need to be replaced each year. Many of these will be secondhand from India or from South Africa. Based on these assumptions, the estimated cost of replacing rolling stock will average about $80 million per year, equivalent to about 0.4 c/ntkm or pkm. Allowing an estimated $20 million for facilities and maintenance equipment, the steady-state investment in the network north of South Africa should therefore be around $200 million per year. The backlog investment is much larger. Assuming a 15-year backlog, an additional investment of about $3 billion would be required ($200 million per year over the 15 years). This expenditure could be spread over a 10-year period, equivalent to an annual cost of $300 million. The com- bined annual cost of rehabilitating and replacing rail infrastructure would therefore be $500 million over 10 years, after which investment would reduce to the steady-state level of $200 million per year. Many proposed new routes would compete with existing road and rail routes. The rates that could be charged by lines running parallel with roads would be limited by the competing road freight rates, typically to 5 c/ntkm at most. For export mineral traffic, the rate that can be charged is also limited by the world-delivered market price, usually around 2­3 c/ntkm. As a consequence, it has been estimated that rail investment would be justified in purely commercial terms only if the forecasted traffic volumes were at least 2 to 4 million tonnes per year, though if they were only expected to cover their operating costs, they could probably be operated successfully at lower volumes of 0.5 to 1.0 million tonnes per year. The Market Railways in Africa are predominantly freight railways. Because of increas- ing competition with the road sector, they are experiencing slow--and sometimes negative--growth. 92 Africa's Transport Infrastructure Some General Characteristics Most railways in Africa carry far more freight than passengers--on average four times as much in traffic units. That ratio continues to increase: only on railways with limited competition from roads does passenger traffic constitute more than 20 percent of traffic units (as shown in figure 3.1 for TRC [Tanzania], Tazara, and Transrail prior to 2005). Figure 3.1 shows that the aggregate ratios for the concessioned and the unconcessioned rail- ways do not differ widely and highlights the small scale of most railways in Africa. Excluding the South African railways, the busier railways typi- cally carry 1 billion TUs per year; Transnet Freight Rail carries this volume in three days. (Other general traffic data can be found in appendix 3c.) The average rail haul in the region is relatively long in the context of overall network size but not especially long compared to that of roads (figure 3.2). Some railways predominantly carry traffic from one end of the system to the other. For example, TRC, Tazara, and Transrail carry freight an average of 1,000 km. On the other hand, some smaller rail- ways--such as the Mozambique and Uganda lines--feed freight to other railways, which subsequently carry traffic a few hundred kilometers farther. South Africa at present accounts for over 70 percent of passenger-km in Africa, mainly because of its suburban commuter business. On most African railways, passenger trips primarily comprise travel between a country's capital city and major provincial centers. The average distances for passenger trips shown in figure 3.2 are therefore primarily based on these services. The Sitarail, Transrail, and Tazara networks have the only significant cross-border passenger flows. (Detailed passenger traffic data can be found in appendix 3c.) Traffic Trends Between 1995 and 2005, gross domestic product (GDP) growth in Africa averaged 4 percent per year. Trade and per capita GDP grew by about 1.5 percent per year. Countries that avoided political upheaval-- such as Mali, Mozambique, and Tanzania--grew up to 50 percent faster than their neighbors. Despite the generally favorable economic background, however, only five railways saw an increase in both passenger and freight traffic. Two--in Namibia and South Africa--were parastatals. Three-- Gabon, KRC, and Central East African Railways Company (CEAR)--were concessioned; the last of these grew despite cyclone damage. Outside of this group, only Botswana experienced an increase in passenger traffic. Railways have thus generally failed to capture traffic despite economic growth. Figure 3.1 Passenger and Freight Traffic of Railways in Sub-Saharan Africa (Annual Average, 1995­2005, Excluding Spoornet) 2,000 1,800 1,600 1,400 traffic units (millions) 1,200 1,000 800 600 400 200 0 Su a T n ts a m Sw ana . (S nd Co C ) ng FM Gh p. Et ana . R B ia . (C in Ni K) ria E Ke n Ta nya Ca nia Si ail Za il Tr bia Ug ail Be a Na ra M la ad i l M law ai C ra i Bo azar d FC da bo ib op ca Re ep en i FM NC r sr ar ge an ep ila a m ta m m w Ga an a nz hi o, . R az Na m De De o, o, ng ng Co Co state-owned railwaysa concessioned railways passenger-km net tonne-km Source: Bullock 2009. Note: CFM = Caminhos de Ferro de Moçambique; CFMK = Chemins de Fer Matadi Kinshasa; FCE = Fianarantsoa Côte Est Railway; SNCC = Société Nationale de Chemins de Fer Congolais. 93 a. Figures drawn from Bullock (2009) follow his distinction between systems that have been concessioned and those that remain in government ownership and operation. Figure 3.2 Annual Average Distance Traveled on Railways in Sub-Saharan Africa, 1995­2005 94 1,600 1,400 1,200 average distance (km) 1,000 800 600 400 200 0 Su ia T n De S tsw a m wa ana . (S d Co C ) ng FM Gh p. Et ana . R B ia . (C in Ni K) ria E Ke n Ta nya Ca nia Si ail Za il Tr bia Ug rail da Na ra M la M wi l ai C ra r FC da bo ep lan ib op ca Re ep en Bo aza i FM NC r ar a ge Be an a m ta s m al m Ga an nz ad . R zi hi o, Na m De o, o, ng ng Co Co state-owned railwaysa concessioned railways passenger freight Source: Bullock 2009. Note: CFM = Caminhos de Ferro de Moçambique; CFMK = Chemins de Fer Matadi Kinshasa; FCE = Fianarantsoa Côte Est Railway; SNCC = Société Nationale de Chemins de Fer Congolais. a. Figures drawn from Bullock (2009) follow his distinction between systems that have been concessioned and those that remain in government ownership and operation. Railways: Not Pulling Their Weight 95 In some cases, war and natural disasters significantly affected railway traffic. For example, Sitarail and the Democratic Republic of Congo's railways both experienced sharp reductions in traffic during periods of civil war. And CEAR suffered badly when a cyclone destroyed a major bridge; it took over two years to find funding for its repair. TRC (Tanzania) was badly hurt by a cyclone, too, in 1997. In other cases, traf- fic was limited by the availability of rolling stock, particularly locomo- tives across many railways.4 When railways have improved this situation by obtaining new or secondhand locomotives or rehabilitating old loco- motives, traffic has increased accordingly. Similarly, infrastructure reha- bilitation on both Madarail and on the Sena line (part of the Companhia dos Caminhos de Ferro da Beira [CCFB] concession) has resulted in a sharp increase in traffic from a low base. Traffic trends over the decade (figure 3.3) were therefore determined more by supply factors than by underlying demand.5 Unfortunately, reducing traffic does not necessarily reduce the need for rehabilitation. Since the downturn in the world economy in late 2008, traffic for some commodities--such as transit cement and rice carried to Burkina Faso by Sitarail and wood exported through Camrail--has dropped 20 to 30 percent on a number of railways in the region, but the need for rehabilitation remains. Passenger Traffic: Limited Prospects Passenger rail services worldwide serve two distinct functions: · Regional and long-distance intercity transport linking major centers to rural areas · Transport of suburban passengers. In many countries in Africa, railways have historically been the only practical mode of intercity passenger transport. In rare cases, rail is still faster than bus (for example, Yaoundé­Ngaoundere in Cameroon and Cuamba­Nampula in northern Mozambique), especially where unpaved roads present difficulties for road traffic in the rainy season. There are generally two passenger classes available, usually called first and third; the overwhelming majority of passengers travels third class (80 to 90 per- cent). Load factors on many trains are often quite high:6 in Tanzania, the average third-class load factor was around 70 percent during the period 1995­2005. Passenger services also carry parcels and small freight, which can increase revenues by about 25 percent. Figure 3.3 Traffic Growth on Railways in Sub-Saharan Africa Annual average, 2001­05, compared with annual average 1995­2000 96 180 160 140 ratio of average 2001­05 to average 1995­2000 (%) 120 100 80 60 40 20 0 Su ia Ta n a m wa na . (S nd Co C ) ng FM Gh p. Et na . R B ia . (C in Ni K) ria E Ke n Ta nya Ca ia Si ail Za ail an a Ug ail da Na ra M la ad i l M law ai C Bo zar Tr bi FC da bo ib op an ca Re ep en i FM NC r r sr ar De S swa a ge Be an ep ila m ta m m Ga a nz hi o, .R z Na t m De o, o, ng ng Co Co state-owned railwaysa concessioned railways passenger freight Source: Bullock 2009. Note: CFM = Caminhos de Ferro de Moçambique; CFMK = Chemins de Fer Matadi Kinshasa; FCE = Fianarantsoa Côte Est Railway; SNCC = Société Nationale de Chemins de Fer Congolais. No comparable data are available for FCE, Transrail, and Madarail. Some railways in the "concessioned" group were only concessioned close to or after 2005. a. Figures drawn from Bullock (2009) follow his distinction between systems that have been concessioned and those that remain in government ownership and operation. Railways: Not Pulling Their Weight 97 Local trains serving villages with no road connections pose a different problem. In Malawi, for example, local services are primarily used by traders bringing goods to and from regional centers. Passenger trains are actually mixed trains, and accompanied goods are loaded into two or three large open wagons. This is a highly inefficient way to bring goods to market: although the trains are well loaded, the revenues barely cover 30 percent of the extra cost incurred in their provision (the "avoidable cost"). Developing feeder roads that provide basic motorized access would usually provide a better long-term solution by lowering the cost of transporting goods and by generally improving access. The suburban rail story is rather different. At the moment there is lit- tle activity outside the services of the South African Rail Commuter Corporation (SARCC). The only other suburban railway in Africa that runs throughout the day is the Petit Train Bleu in Dakar, which serves only a single route. Most suburban lines are locomotive-hauled commuter railways with few, if any, services outside peak periods. While some cities in other parts of Africa also operate commuter services, with the excep- tion of Dakar, such services are generally limited to one or two peak hours and go only a short way out along the main line. Some cities, however, have plans to introduce modern commuter net- works. The most notable is Lagos, which is currently seeking private oper- ators for two new urban rail lines that together are expected to carry nearly 2 million passengers per day. This is more than the total for all existing passenger services in South Africa. The problem is that the financial basis for such projects is typically weak, as a consequence of governments trying to maintain low fares with- out paying any direct subsidy. Even in South Africa, more than a third of the rolling stock in the urban rail network is forecast to go out of service within the next three years. According to experience in most large conur- bations in the world, urban passenger railways require substantial exter- nal funding for capital and recurrent operating costs. Moreover, it is desirable that they should be operated by companies separate from the existing railway (as is the case of the South African Rail Commuter Corporation), and preferable that they have the support of a multimodal conurbation transport authority (such as the Lagos Metropolitan Area Transport Authority). Passenger Fares: A Constraint on Financial Viability Governments typically set fares at what they consider to be affordable levels, usually ranging from 1 to 3 cents per passenger-km (c/pkm). This 98 Africa's Transport Infrastructure is almost invariably below cost. In a few cases, the government has for- mally declared the carriage of passengers at such low rates to be a "pub- lic service obligation" for which companies should be directly compensated by the government itself. Unfortunately, the compensation is frequently inadequate to cover the losses imposed or is not paid in a timely manner. Railways must then attempt to use freight revenues to cover the cost of operating passenger services (see appendix 3d for more details). The difficulty is compounded by the fact that road networks have improved, so that in many corridors, buses and shared taxis compete with railways in terms of both price and service frequency. Fares reflect service quality; vehicles range from luxury buses on a few routes to ordinary buses, minibuses, and converted freight vehicles. Bus fares are typically about 30 to 50 percent higher than the economy rail fare. On the other hand, buses are up to twice as fast as railways (figure 3.4) and generally run much more frequently. The long-term prospects for interurban rail passenger services are therefore poor. The cost of maintaining rail track and signaling to even marginally compete with passenger travel on an average sealed road-- which takes place at around 70 km per hour (km/hr)--is significantly more than maintaining track and signaling for freight--which travels at 30 to 40 km/hr. To justify the substantial cost of constructing new, medium-speed (200 km/hr) interurban railways, corridors would need to show the potential for substantial demand (several million passengers per year) by people able to pay higher rates. Few, if any, corridors in Africa meet these requirements in the medium term. Freight Traffic: Dominated by Bulk Movements to and from Ports Bulk and semibulk commodities, primarily to and from ports, dominate freight traffic on railways in Africa. Of the railways included in figure 3.5, only Botswana (Botswana Railway, BR) has a large proportion of nonport traffic, most of it outgoing raw materials or incoming products (such as cement and petroleum) from South African manufacturers. The commodities transported by rail reflect the economic structure of countries: mining products (copper, tin, manganese, stone, and coal) are important in Gabon, South Africa, and Zambia; and timber and export crops (cocoa, coffee, cotton, cereals) are important in West Africa. (More details on freight traffic can be found in appendix 3c.) Import traffic generally exceeds export traffic. The only exceptions are a few railways--such as in Gabon--that have substantial export mineral Railways: Not Pulling Their Weight 99 Figure 3.4 Comparison of Bus and Rail Fares and Travel Times 0.8 Dar to Mwanza 0.7 Douala to Yaounde Nairobi to Mombasa bus travel time: rail travel time 0.6 Ouagoudougou to Bobo Dar to Kapiri Mposhi 0.5 Gabarone to Dar to Dodoma Francistown 0.4 Windhoek to 0.3 Swakopmund 0.2 0.1 0.0 0.0 0.5 1.0 1.5 2.0 2.5 bus fare: rail fare Source: Bullock 2009. flows. International traffic tends to dominate railways that cross interna- tional borders. The main exception is the KRC, which transports much more traffic to and from Nairobi and other centers than to and from Uganda. Imports are primarily manufactured products such as cement, petroleum products, and general freight. Higher-value cash crops (such as coffee from Uganda) increasingly travel in containers, particularly on routes that cross a national border. Severe directional imbalances in traffic are the norm. Even when the tonnage transported is nearly the same in both directions, many com- modities require specialized wagons so that trains are rarely fully loaded in both directions. In some cases (such as export traffic from Zambia to connected ports), road vehicles delivering imports tend to backload freight at a marginal cost, leaving railways to transport the remaining freight without a compensating return load. Such practices accentuate the imbalance in rail traffic. Rail traffic has decreased in many countries over the past years, coin- ciding with the abolition or restructuring of statutory agricultural market- ing organizations. These organizations were often the only means for producers to market their crops, as they provided depots at key points on rail networks where producers could bring their products for storage and 100 Africa's Transport Infrastructure Figure 3.5 Commodities Carried on Select Railways in Sub-Saharan Africa 100 80 60 percent 40 20 0 C l n il l E C C AR a B l BR i ai ai ra ra bi CD CF bo GR UR TR sr ar CE m ta m /C Ga an ad Si Ca Za DN Tr M /C M CF railway timber cement fertilizer petroleum ores and minerals agriculture general Source: Bullock 2009. Note: CDE = Chemin de Fer Djibouti-Ethiopien; CDN = Corredor de Desenvolvimento do Norte; CFM = Caminhos de Ferro de Moçambique; GRC = Ghana Railway Company; URC = Uganda Railways Corporation. The data relate to specific years or averages of years around 2001. As the proportions do not change much from year to year, the broad picture shown is valid despite the fact that the data do not refer to the same year for each railway. subsequent dispatch. Marketing channels for agricultural products are now more diversified, and, as a result, the railways have steadily lost mar- ket share. Abandoned rail-connected warehouses for export cash crops at ports such as Dar es Salaam are testimony to these changes. Inland distribution networks for consumer and intermediate products have similarly changed. Although there are still inland depots for petro- leum products, direct deliveries from main depots and refineries to end users are now more common, with small consignment sizes that are far better suited to road transport. General freight, whether containerized or not, is dispatched in relatively small consignments; mixed loads, with freight from two or three suppliers to the same destination, are common. Factor productivity is low (see appendix 3e). For traffic of this type, the costs of pickup and delivery can also make rail transport prohibitively expensive. In many cases, this small general freight and mixed-load traf- fic was lost to roads and has not been recovered even as railways were concessioned. Railways: Not Pulling Their Weight 101 Freight Tariffs: Increasingly Competitive Average freight tariffs between 1995 and 2005 were typically 3­5 c/ntkm (figure 3.6; appendix 3d).7 Railways originally based tariffs on the value of the commodity being transported, charging low rates for low-value commodities such as fertilizers and high rates for manufactured goods.8 Tariffs were sometimes affected by policy support for particular sectors, such as agriculture, often gained by special-interest lobbying. Nowadays, tariffs are more determined by demand, being limited by competition either from roads or alternate routes, except in the occa- sional case of semi-monopolies--such as the Société Transgabonnaise (SETRAG) in Gabon. Railways with little competition, including the Democratic Republic of Congo railways (Société Nationale de Chemins de Fer Congolais [SNCC] and Chemins de Fer Matadi Kinshasa [CFMK]), Chemin de fer Congo-Océan (CFCO) in the Republic of Congo--for which road competition lacks security and is expensive (or impossible)--tend to have higher rates (see figure 3.6). But many rail- ways do not fully understand their own cost structure, and their response to road competition has therefore been imperfect. Other factors also affect tariff structures. For example, tariffs in Uganda's network are distorted by rail ferry operations, the short length of port-access lines, and a deliberate policy of equalizing tariffs across all three routes (direct rail, ferry via Kenya, and ferry via Tanzania) to pro- mote competition. Despite the pressure from road competition, tariffs still vary substan- tially both among commodities and among countries (figure 3.7). Tariffs for petroleum products and container traffic are generally high, while those for agricultural products and semibulk commodities such as cement and fertilizer are low. These differences reflect not only traditional, value- based tariff structures but also relative costs of carriage, volume (for example, many railways negotiate contract rates with high-volume users), and traffic direction. Thus, bulk commodities, which have higher net loads per wagon, are cheaper to carry than petroleum, which is normally car- ried in tank wagons that have a comparatively high ratio of gross to net tonnes and are almost always returned empty. Rates in the low-volume direction are generally around half to two-thirds of those in the high- volume direction, with rail rates similarly discounted under road rates. The economic viability of transporting a specific commodity by rail is significantly affected by whether the origin and destination are rail connected--that is, whether a mine or a cement works has a rail siding Figure 3.6 Average Tariffs for Passenger and Freight Traffic, 1995­2005 102 0.20 0.18 0.16 yield (US$ per pkm or ntkm) 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0 Su ia T n t ra m w na . (S nd Co C ) ng FM Gh . Et ana . R B ia . (C in Ni ) ria E Ke n nz a Ca nia Si il Za ail Tr bia Ug rail da Na ra M la M awi l p ai C K ra Ta ny FC da bo ib op ca Re ep en Bo aza i FM NC r ar De S swa ge Be an ep ila a m ta s m al m Ga an ad hi o, . R az Na m De o, o, ng ng Co Co a state-owned railways concessioned railways passenger freight Source: Bullock 2009. Note: CFM = Caminhos de Ferro de Moçambique; FCE = Fianarantsoa Côte Est Railway. No data are available for FCE, Nigeria, and Swaziland. a. Figures drawn from Bullock (2009) follow his distinction between systems that have been concessioned and those that remain in government ownership and operation. Figure 3.7 Average Tariffs by Commodity for Select Railways in Sub-Saharan Africa 10.0 9.0 8.0 7.0 US cents per ntkm 6.0 5.0 4.0 3.0 2.0 1.0 0 petroleum containers cement fertilizer cereals cotton products commodity Sitarail Camrail Transrail Madarail TRC 2001 ZR 2003 CEAR 2003 Source: Bullock 2009. 103 Note: ZR = Zambia Rail; Where not shown, average yields are for the period 1995­2005. 104 Africa's Transport Infrastructure and ready access to a port or power station. If the origin and destination are not well connected, road shipment to and from the railway can cost up to the equivalent of 200 to 300 km of rail transport. Demand must be adequate for the construction of new rail siding to be economical. Traffic that comes from multiple origins and must be collected at a central depot before being dispatched by rail is therefore most vulnerable to road com- petition. On the other hand, mineral and other bulk loads from a single source tend to use rail as long as sufficient service capacity is available. Bulk loads are not immune to road competition, however; roads have been used for relatively short-distance intermine traffic in the Zambian Copper Belt, even though a rail network built for that purpose connects the mines and processing plants. A cross-country comparison of freight rates and market shares needs to take into account not only physical factors such as infrastructure, vehicle type and quality, and freight, but also the direction of travel, overloading, and other institutional factors (figure 3.8). The wide variation in freight rates among geographical areas reflects regional variations in infrastruc- ture and road vehicles, average length of haul, and institutional factors such as unofficial en route charges (bribes or forced extraction of illegal tolls), border-crossing procedures, and the impact of the freight associa- tions common in Central and West Africa. Some of the reasons for the regional variations are straightforward: the poor condition of roads in Central Africa and the Central Corridor in East Africa connecting Dar es Salaam, Burundi, and Rwanda; the very large trucks operating throughout southern Africa;9 and the impact of the freight associations. Why Are Railways Uncompetitive? The inability to attract more traffic despite a large price advantage can be explained in several ways. First, there is the extra cost of local road pickup and delivery for long-distance rail freight transport. Second, service quality (transit time, reliability, and security) is generally poor. Rail infrastructure is below par along most corridors. For a typical corridor, rail rates need to be about 15 percent less than road rates to cover the additional cost of road access to rails, and a further 15 percent less to compensate for infe- rior service quality. The main obstacle to competitive rail service quality is institutional-- a lack of trade facilitation and cross-border coordination. In the main Railways: Not Pulling Their Weight 105 Figure 3.8 Indicative Freight Rates: 12-meter Container Inland from Port, 2003 5 4 US$/container-km 3 2 1 0 ca ca ) ) ) a NC CC SC ric fri fri a( a( Af a( tA lA ric ric ric rn ra es Af Af he Af nt W st Ce ut st st Ea Ea Ea So region road rail Source: UNCTAD 2003. Note: CC = central corridor; NC = northern corridor; SC = southern corridor. north-south corridor from South Africa to the Democratic Republic of Congo and Tanzania, rail transit times from the Democratic Republic of Congo to Durban have been quoted as 38 days--9 days for travel and 29 days for interchange and border crossing, despite the fact that the rail corridor is effectively under the control of a single operator from the border of the Democratic Republic of Congo to South Africa. By way of comparison, competing road transport by truck reportedly takes 8 days overall, only 4 of which are at border crossings. Over and above these external disadvantages, many state-owned rail- ways have difficulty competing against road operators because they do not have the freedom to set rates according to demand. Competition from the road sector is strongest in southern Africa, which has the most liberal market structure, the largest trucks, and the best roads. Two other major factors influence road competitiveness: user charges and the prevalence of overloading. Few governments charge trucks adequate road use fees (see chapter 2 of this book), and overloading of trucks is 106 Africa's Transport Infrastructure commonplace, increasing necessary maintenance costs. Requiring railways to fund 100 percent of rail maintenance and improvements while tolerat- ing inadequate road use fees and vehicle overloading on arterial routes cre- ates a handicap almost impossible for most general freight railways to overcome. Many individual African railways also face competition in the freight market from other corridors (Mbangala 2001). In West Africa, the inland countries (Burkina Faso, Mali, and Niger) have a choice of ports to and from which they can transport goods: Dakar (Senegal), Abidjan (Côte d'Ivoire), Conakry (Guinea), Takoradi and Tema (Ghana, by road), Lomé (Togo), and Cotonou (Benin). The Great Lakes region in East Africa has a similar range of competing outlets to the sea, only some of which are served by rail. In sum, while most railways are able to carry bulk minerals with rea- sonable efficiency, they must offer a reasonable level of general freight service if they are to compete with roads. Conventional state-owned rail- ways are poorly equipped to provide door-to-door service because of their fixed rates, low service levels, lack of commercial incentive to change, and the conservative management behavior that usually goes with these characteristics. Concessioning can help. Concessionaires have already shown that they are prepared to use a range of initiatives to improve service quality and compete with roads. Some are physical, such as Sitarail's proposal to con- struct an intermodal terminal in Ouagadougou to service the surrounding region. Others are procedural, such as Zambia Rail's introduction of com- pany customs bonds to reduce waiting times for import traffic at Victoria Falls.10 Above all, however, private concessionaires have the commercial freedom, flexibility, and incentive to provide services that meet demand. Concessioned railways worldwide are increasingly integrated with trans- port chains either through participation in third-party logistics systems or by direct connection to primary production processes. Institutional Arrangements Until the 1980s, almost all African railway companies were publicly owned corporations, subject to general supervision by a ministry of trans- port mandated to develop and implement policy. Some had French-style contract plans that aimed to explicitly define the relationship between governments and railway companies. These arrangements were generally Railways: Not Pulling Their Weight 107 ineffective because governments rarely met their formal obligations to the public corporations' management. A further step toward financial and managerial autonomy was the introduction of management contracts, under which an independent specialist agency agrees to manage publicly owned assets to achieve specified objectives. For example, the Indian RITES company sup- plied senior management to Zambia in the 1980s and Botswana Railways in the 1990s, and had a full management contract with Nigerian Railways in 1979­82. Togo Railways was managed by CANAC for some years but is now managed by RITES in association with a local cement company. The Democratic Republic of Congo has attempted two such contracts: the first, known as Sizarail, ceased with a change in government in 1997; the second is still in operation. The weakness of these arrangements was that so long as government was responsible for the provision and financing of all physical assets, the operational scope of the management contractor was severely limited. For more effective commercial rail management, emphasis has been shifted to the creation of railway concessions since the early 1990s. In these concessions, the state remains the owner of some assets (typically infrastructure) but transfers the others (typically the rolling stock) to a concessionaire. The concessionaire assumes responsibility for operating and maintaining the railway. In some cases, such as the Sitarail concession, the government also purchases new rolling stock, which the concessionaire then finances with annual payments to the government. The first concessioned railways were in West Africa: the Abidjan- Ouagadougou railway linking Côte d'Ivoire and Burkina Faso was conces- sioned in 1995, followed by railways in Cameroon, Gabon, and Malawi at the end of the 1990s. With the exception of southern Africa (Botswana, Namibia, South Africa, and Swaziland), which has not yet faced the financial crises that precipitated reform in most other countries, and countries suffering or recovering from civil disruption (Angola, Democratic Republic of Congo, and Zimbabwe), most countries in Africa are in the midst of some type of railway reform. Of the 30 countries with state-owned railways, 14 have opted for a concession arrangement, often under the pressure of multilateral and bilateral organizations that prom- ise to finance asset rehabilitation and renewal. The railway in one coun- try (Democratic Republic of Congo) operates under a management contract. Another four countries have begun the concession process, and others are planning to do so (map 3.2; appendix 3a). 108 Africa's Transport Infrastructure Map 3.2 Railway Concessions Awarded in Africa since 1990 Sizarail DRC Transrail 1995/1997/2008 Senegal/Mali 2003 RVRC Kenya/Uganda Sitarail 2006 Côte d'Ivoire/ Burkina Faso 1995 TRC Canac/WACEM Tanzania Togo 2007 1995/2002 Camrail CEAR Cameroon Malawi Transgabonais 1999 1999 Gabon 1999 Madarail Railway operated by state RSZ Madagascar railway company Zambia 2003 2003 PSP project planned or CDN underway Mozambique 2005 Part of rail network now BBR CCFB Zimbabwe Ressano Garcia Mozambique under private management Mozambique 1997 2005 Cancelled Railway now under private management Source: Bullock 2009. Note: BBR = Belt Bridge Railway; CDN = Corredor de Desenvolvimento do Norte; RSZ = Railway Systems of Zambia; RVRC = Rift Valley Railways Consortium; WACEM = West African Cement Company. Three different dates are given for Sitarail because the concession was relet on three occasions. Governance and Management of State-Owned Railways Most of the remaining state-owned railways are subject to significant political and governmental influence. Arrangements vary across coun- tries, but typically the sector ministry (normally transport) exercises political and administrative control, while the ministry of finance exer- cises financial control. Boards generally comprise a combination of ministry and internal senior management officials--themselves often appointed by the government--with occasional staff representation. Parliaments provide nominal oversight. All too frequently, however, this oversight is limited to an audit of company accounts presented in an annual report, sometimes several years after the year in question. Parliamentary sessions are often too short for the detailed review that effective control would require. The governing regulatory frameworks generally grant financial and managerial autonomy to state-owned railways, and management methods are supposed to be similar to those of private businesses. At the same Railways: Not Pulling Their Weight 109 time, however, legal and regulatory frameworks provide greater opportu- nities for state intervention--at both the institutional and jurisdictional levels--than would be the case for privatized railways. Railway commercial initiatives are subject to frequent political interference, and government- authorized representatives in companies have decision-making capabili- ties. This latent conflict between the control and decision functions in state-owned railways discourages effective management. Governments and politicians also have ulterior motives--often dic- tated by mutually exclusive social, electoral, and economic interests-- which further complicate both the management of state-owned companies and the evaluation of their performance. Management often focuses on merely breaking even on a cash basis, which almost inevitably leads to financial difficulties when asset renewals are due. Structure of Concessions Railway concessions vary substantially in terms of contract length, the range of assets transferred to the concessionaire, the attribution of respon- sibility for investment during the concession, and the limitations on the commercial freedom of concessionaires, particularly in regard to passen- ger services.11 The concession contracts in Africa to date are summarized in table 3.1. Few governments have seriously considered the European model of full vertical separation in which track management and train operations are performed by different companies. There is thus little scope for exten- sive competition between different companies operating on the same track. But there are a number of cases of independent, noncompeting companies running trains on state-owned or concessioned railway lines. Magadi Soda Works ran its trains to Mombasa over the KRC line in Kenya and continues to do so over the concessioned Rift Valley Railways line. Senegal's concession to Transrail excluded the Dakar suburban service and traffic from the Société d'Exploitation Ferroviaire des Industries Chimiques du Sénégal (SEFICS), both of which now pay track charges to use Transrail track. In a rather more complex case, the concession offering in Zambia, which has extensive intermine operations in the Copper Belt, allowed bidders to choose to include any combination of Zambia's three railways: mainline operations, intermine operations, and passenger serv- ices. The winning bidder initially chose to include all three but subse- quently decided to include substantially fewer intermine services in the concession. Those not included in the concession are still able to run trains over the concessioned tracks. 110 Table 3.1 Key Features of Concessions, 1993­2008 Initial capital Planned 5-year Date of Initial duration of of concession investment Public service Country Concessionaire contract contract (years) (US$ millions) (US$ millions) obligation West Africa Burkina Faso Sitarail 1995 15 8.8 63.3 Yes, but renegotiated Côte d'Ivoire n.a. n.a. n.a. n.a. n.a. n.a. Mali Transrail 2003 25 17.2 55.4 No Senegal n.a. n.a. n.a. n.a. n.a. n.a. Togo CANACa 1995 n.a. n.a. n.a. No WACEMa 2002 25 n.a. n.a. n.a. Central Africa Cameroon Camrail 1999 20 18.5 89.6 Yes Democratic Republic Sizaraila 1995 n.a. n.a. n.a. n.a. of Congo Vecturisa 2008 2 n.a. n.a. n.a. Gabon Trans-Gabonais 1999 20 n.a. n.a. n.a. SETRAG 2005 30 n.a. 157 Yes East Africa Kenya RVRC 2006 25 n.a. About 100 Yes Uganda n.a. n.a. n.a. n.a n.a. n.a. Tanzania TRC 2007 25 n.a. 88 Yes Southern Africa Malawi CEAR 1999b 20 n.a. n.a. Yes Mozambique CCFB 2004 25 19.7 152.5 Yes Mozambique CDN 2005 15 n.a. n.a. Yes Madagascar Madarail 2003 25 5.0 36.1 Yes Zambia RSZ 2002 20 6.1 14.8 Yes Zimbabwe NLPI 1998 30 n.a. 85c No Sources: Data collected from companies; Pozzo di Borgo 2006. Note: CDN = Corredor de Desenvolvimento do Norte; RSZ = Railway Systems of Zambia; RVRC = Rift Valley Railways Consortium; SETRAG = Société Transgabonnaise; WACEM = West Africa Cement Company; n.a. = not applicable. a. Management contract. b. Transfer finally occurred in 2005. c. Reported construction cost. 111 112 Africa's Transport Infrastructure There is usually specific provision for the financing of system rehabil- itation at the beginning of a concession.12 In addition, concessionaires are normally responsible for financing track and rolling-stock maintenance and renewal during the period of the concession. In some cases, concession- aires have received loans to finance rolling stock, but many low-volume operators use secondhand equipment instead. Railway concessions in Africa generally rely on either of two models for financing initial infrastructure investment: · Governments finance the initial track rehabilitation and renewal costs, generally with specific-purpose loans from international financial institutions (IFIs), which offer grace periods, lengthy loan tenors, and below-market interest rates. · Governments do not finance initial track renewal but commit to com- pensate concessionaires for their investment by the end of the conces- sion agreement (as in the case of the KRC/Uganda Railways Corporation [URC], TRC, and Zambia railways). Special cases are the Beitbridge Railway, which relies on take-or-pay clauses (which give it a guaranteed revenue against which it can borrow); and Nacala, which is being funded at semicommercial rates. In both models, the government usually agrees to purchase the unamortized portion of any infrastructure investment that the conces- sionaire will have financed by the end of its contract. When conces- sionaires fear that governments might not be willing or able to make such payments, they might limit their infrastructure investments as the contract period nears. Kenya and Uganda solved this problem in the KRC/URC concession by obtaining a partial risk guarantee from the World Bank to securitize their payment obligations to the conces- sionaire. In the most common arrangement, the state remains the owner of some or all of the railway's assets (normally the infrastruc- ture) and transfers responsibility for operations, risks, and expenses to the concessionaire in the concession agreement. Regulatory Framework for Concessions The introduction of concessions has necessitated substantial changes in railways' legal and regulatory frameworks. Such changes were particularly important in Anglophone countries, where railways were state owned and the responsible ministry imposed wide-ranging economic regulations while leaving railways to self-regulate safety.13 Concessions require more Railways: Not Pulling Their Weight 113 transparent economic and safety regulations. Tanzania passed a new rail- way law in 2003, and Zambia has drafted a new law that has not yet been implemented. Although Malawi plans to amend its railway act, it has still not done so. Concessions agreements almost always have a predetermined dura- tion. The concessionaire therefore generally leases the infrastructure (and sometimes the rolling stock) from the government. This arrange- ment requires that the ownership of the assets be transferred from the existing state railway operating company to a successor asset ownership authority. While the specific arrangements vary, such bodies generally are responsible for ensuring that the concessionaire maintains the railway assets properly and for funding capital expenditure that is not the con- cessionaire's responsibility. They often also effectively regulate safety, as they award operating licenses on the basis of technical competence. In Francophone Africa, these are the "patrimony organizations" (legacy organizations such as the Société Ivoirienne de Patrimoine Ferroviaire in Côte d'Ivoire and the Société de Gestion du Patrimoine Ferroviaire du Burkina in Burkina Faso). Agencies were also established with similar functions in some Anglophone countries (such as the Reli Assets Holding Company in Tanzania). The new bodies are often in theory funded from concession fees, but failure to make adequate provision for them is an ongoing problem in some countries and severely handicaps their ability to provide effective regulation. Contracts normally distinguish between freight tariffs, which are gen- erally deregulated, and passenger tariffs, which the state tends to control. Regulated passenger services are often managed using agreed-upon schemes under which operators are eligible for financial compensation (public service obligations payments) whenever the regulated tariffs do not cover their operating costs. These schemes have often failed to pro- tect private operators, however, as governments have not honored their subsidy commitments. Passenger tariff indexation, when applied (such as for Sitarail, CEAR, and Transrail), is triggered by changes in conventional inflation indexes. Most concession contracts prevent rail operators from using promotional tariffs for more than a year if they do not cover their operating costs. Economic regulation is frequently left to the general powers of a com- petent competition commission. In Zambia, for example, the Competition and Fair Trading Act, as administered by the Zambia Competition Commission, has broad powers of referral for the abuse of market power by dominant suppliers. In Tanzania, a regulatory body established specifically 114 Africa's Transport Infrastructure for the transport sector (Surface and Marine Transport Regulatory Authority, or SUMATRA) has authority over all land and marine transport. The sepa- rate railway regulatory agencies of Mali and Senegal have merged into a common railway monitoring agency serving both countries. Concession agreements generally delegate powers of referral for tariffs and control over third-party access to either the government or an independent authority.14 In reality, however, many railway concessions in Africa lack formal regula- tory structures with real power and are thus susceptible to market abuse. Despite the lack of effective regulatory agencies, railways in the region are unlikely to require frequent protection from market abuse because the road transport sector, which generally offers services competitive with rail, limits the market power of concessions. Where freight rates have increased following concessioning, there has generally been a correspon- ding improvement in service quality. In most cases, rail freight rates are effectively determined by competition, either from roads or from rail routes serving an alternate port (such as the ports of Beira for the Nacala corridor and Abidjan, Lomé, and adjacent ports for Transrail). Relatively few railways have true monopolies over freight (Trans-gabonais's trans- port of bulk minerals such as manganese is a rare example).15 A detailed study of four transport corridors involving a railway concession (and in some cases an associated port) confirmed the general absence of market abuse by monopolies (Pozzo di Borgo 2006). If a concessionaire fails to comply with the terms of a concession, there are normally procedures for terminating the concession. To date, however, only three concessions (Ressano Garcia, which never became operational, Trans-Gabonais, and the RITES contract in Tanzania) have been termi- nated,16 while two concessions (Transrail and Rift Valley) have changed operators. Concessions have not been without difficulties. Concessionaires have faced delays and disputes regarding government compensation for unprofitable services. Other conflicts have centered on concession fees, time frames, and staff no longer required following concessioning. The failure of government ministries to coordinate their actions--which have included administratively imposed salary increases, restrictions on access to container facilities, and unfunded public service requirements--has also negatively affected the performance of several concessions. Concessionaires: Motivated by Broader Self-Interest Rail concessions in Africa have attracted a limited pool of private opera- tors, mainly from southern Africa or from outside Africa completely. Railways: Not Pulling Their Weight 115 These operators are of two types: those seeking to vertically integrate their distribution chains by acquiring dominant positions in specific pro- duction and transport sectors, and those specializing in a single transport activity (such as railways or ports). The first group appears willing to accept low rates of return from individual components of their distribu- tion chains (especially railways) as long as the control that vertical inte- gration provides yields sufficient benefits overall. The best example of this type of operator is the Bolloré group, which is the largest or second- largest shareholder in several railway and port concessions in Africa and also operates as a freight forwarder. Previously, the group also had agricul- tural production subsidiaries. The most prominent of the second type of operators include Sheltam from South Africa (another South African operator, Comazar, is now defunct), NLPI from Mauritius, and RITES from India. They invest in transport operations, suggesting that concessions can be sufficiently prof- itable to attract private operators. The business cases for their rail invest- ments often appear weak,17 however, suggesting that these companies may be seeking the financial benefits of managing large investment plans (financed for the most part by governments) rather than long-term busi- ness cash flows. Private companies are the majority shareholders of all concessions to date (table 3.2). State participation is highest in Mozambique, which has a 49 percent stake in both the CCFB and Corredor de Desenvolvimento do Norte (CDN) and is also a significant shareholder of the adjacent CEAR concession, and Tanzania, where the government has a 49 percent stake in the TRC. The government of Madagascar owns 25 percent of Madarail, and governments own 10 to 20 percent of Sitarail, Transrail, and Camrail. Local private ownership of any kind within competing consor- tia has been generally limited and in any case appears to make the process more vulnerable to political manipulation. Madarail has the highest level of local private ownership, at 24 percent, compared to 10 to 20 percent for Sitarail, Setrag, and Camrail. Employee shareholding remains under 5 percent where it exists at all. The boards of concessionaires generally reflect shareholding arrangements and thus include government- appointed members. Operational Performance The productivity of labor and rolling stock of railways in Africa is low com- pared to railways elsewhere. This is not surprising given most networks' 116 Africa's Transport Infrastructure Table 3.2 Initial Concession Shareholdings Concessionaire Shareholder Percentage ownership Sitarail SOFIBa 67 Governments of Côte d'Ivoire and Burkina Faso 30 Employees 3 Transrail Canac-Getma (France and Canada) 78 Governments of Mali and Senegal 22 Camrail SCCF (Cameroon)b 85 Government of Cameroon 10 Employees 5 Setrag Comilog (France) 84 Local private operators 16 RVRC Sheltam (South Africa) 61 Other foreign investors 15 Local private investors 25 TRC RITES (India) 51 Government of Tanzania 49 CEAR Edlow Resources and Railroad Development Corporation (United States) 51 Mozambican local investors (including CFM) 49 CCFB Irconc 25 RITES (India) 26 Government of Mozambique (through CFM) 49 CDN Same as CEAR Madarail Madaraild 51 Government of Madagascar 25 Manohisoa Financière 12.5 Other private operators 11.5 RSZ NLPI Majority Transnet (South Africa) Minority Source: Bullock 2009. Note: Ownership in a number of concessions has changed since these data were published. CDN = Corredor de Desenvolvimento do Norte; CFM = Caminhos de Ferro de Moçambique; RSZ = Railway Systems of Zambia; RVRC = Rift Valley Railways Consortium. a. Société Ferroviaire Ivoiro-Burkinabé (SOFIB) was majority controlled by Bolloré (France). Sixteen percent of stock was intended for sale on the Abidjan Stock Exchange. b. Société Camerounaise des Chemins de Fer (SCCF), a holding company controlled by Bolloré. Comazar, a pri- vately operated and managed company that included South Africa's Spoornet and Transurb Consult (a subsidiary of the Belgian National Railways), also held substantial shares of SCCF, but has since sold its interests in Cameroon and elsewhere to Vecturis, a Belgian firm founded by two ex-Comazar employees. c. Indian Railways Construction Corporation. d. Madarail was majority owned by Comazar at the time of concession. low traffic volumes and poor infrastructure conditions (Mbangala and Perelman 1997). Concessioned railroads have better productivity indica- tors, a result explained in part by increases in traffic but mostly by major cuts in employment. Railways: Not Pulling Their Weight 117 Labor Productivity: Low but Improving Most railway companies in the region have streamlined their workforce to some extent over the past 10 to 15 years. This is often a prelude to con- cessioning but in some cases also reflects a company's general effort to improve efficiency. Although the labor productivity of most African rail- ways has improved, it remains low by world standards:18 few railways in the region annually achieve over 500,000 TUs per employee (figure 3.9). South Africa's Transnet Freight Rail has the highest labor productivity of any railway system in the region. Its average productivity was 2.5 million TUs per employee between 1995 and 2005, and it reached 3.3 million in 2005 (Thompson 2007). This figure reflects the intrinsically high produc- tivity of mineral transport: the labor productivity of its dedicated iron ore and coal export lines (Orex and Coalex) were 9 million and 38 million, respectively, compared to only about 1.5 million for its residual general freight business (about 40 percent of Spoornet's traffic). Gabon also had high average labor productivity over that period, reaching 1.8 million TUs per employee in 2005. Like Spoornet, Gabon has a high proportion of mineral traffic; a third party owns and operates the trains used for its mineral transport, which further increases labor productivity. The labor productivity of most other railways in the region is low. In some cases, this deficiency reflects the railways' lack of outsourcing and continued reliance on labor-intensive methods, for example, in track maintenance and wagon loading. The very low productivity of other rail- ways (for example, Benin, Democratic Republic of Congo, Ghana, and Nigeria all achieve less than 100,000 TUs per employee), however, is a result of their failure to cut staff despite declining traffic. When wages are low, redundant employment may not be financially catastrophic for a rail- way. Nevertheless, having too little work for too many employees erodes morale and is a strong disincentive for improving efficiency in the use of other assets. Railways that pay low wages also find it hard to recruit and retain the technically competent staff required by the technology that could improve service quality. Rolling-Stock Productivity The productivity of rolling stock is determined by several factors: the pro- portion of usable stock, the proportion of available stock, the usage of available stock (in hours per day), the commercial speed of operations, and the power of available locomotives. Railways in Africa are deficient in all of these respects. Better management can certainly improve rolling stock and locomotive productivity by disposing of surplus assets and Figure 3.9 Labor Productivity of Railways in Sub-Saharan Africa 118 3.2 labor productivity (million TU/employee) 2.8 2.4 2.0 1.6 1.2 0.8 0.4 0 Na net Su ia T n ts ra . R az a . (S nd Co C ) ng FM Gh p. Et ana . R B ia . (C in Ni K) ria E Ke n nz a Ca nia Si il Za rail Tr bia Ug rail da Na ra M la M wi l ai C ra m w an Ta ny FC da bo ib op ca Re ep en Bo aza i FM NC ar a ge Be an ep ila a m ta s m r al m De S w Ga an oo ad hi o, Sp m De o, o, ng ng Co Co a state-owned railways concessioned railways average 1995­2005 current Source: Bullock 2009. Note: CFM = Caminhos de Ferro de Moçambique. No data were available for FCE (Fianarantsoa Côte Est Railway). Some railways were concessioned close to or after 2005. This figure should therefore not be used to assess the effect of concessioning on productivity. Current figures are for the latest available year, typically 2005. a. Figures drawn from Bullock (2009) follow his distinction between systems that have been concessioned and those that remain in government ownership and operation. Railways: Not Pulling Their Weight 119 improving the supply chain for spare parts, but it cannot improve low axle loads and low commercial speeds. Low fleet availability limits the locomotive productivity of railways in the region. On a given day, less than 40 percent of many railways' stock is available. In 2008, for example, the average availability of SNCC's fleet in the Democratic Republic of Congo was limited to only 10 of 22 main- line electric locomotives and 13 of 47 diesel locomotives, or 45 percent and 28 percent, respectively. At the other extreme, Swaziland, with a labor productivity reaching 2.159 million TUs per employee in 2005, also showed high locomotive productivity because, like Botswana, it carried a substantial proportion of transit traffic (75 percent), which is relatively simple to operate and for which third parties own and maintain the wag- ons. Gabon, which also has high locomotive productivity, and Swaziland are relatively new railways, while Botswana has benefited from substan- tial investment in the past 30 years. Passenger carriage productivity varies widely among railways (figure 3.10). Between 1995 and 2005, for example, Camrail, Ethiopia, Nacala, and Tazara averaged around 5 million pkm per car annually, com- pared to under 500,000 pkm per car in the Democratic Republic of Congo and Sudan. Some railways with high productivities simply operate small fleets of overcrowded carriages. More typical are railways such as Camrail, which operates a regular service with reasonable load factors--say 40 to 50 passengers per vehicle (Murdoch 2005). For these railways, annual dis- tance traveled per vehicle averages 100,000 to 130,000 km with an aver- age load of 40 to 50 passengers; passenger carriage productivity therefore ranges from about 4.0 to 6.5 million pkm. As long as demand is sufficient to keep most of the fleet operating, availability is generally reasonable-- often well over 80 percent. Freight wagon productivity is around 1 million ntkm per wagon in Swaziland (largely due to the prevalence of transit traffic carried in "foreign" wagons, which do not appear in the base for the calculation, while the traffic that the foreign wagons carry in Swaziland is counted). Several other railways have a productivity rate of over 500,000 ntkm per wagon. Yet wagon productivity on many state- owned railways (including some concessioned to private operators at the end of the study period) is very low--less than 200,000 ntkm per wagon (figure 3.11). This is usually the result of demand for rail freight transport falling without corresponding reductions in the wagon fleet (wagons can have useful lives of 50 years or more). Wagon mainte- nance is straightforward, normally requiring only a limited range of Figure 3.10 Passenger Car Productivity (Average, 1995­2005) 120 5.5 5.0 4.5 carriage productivity (million pkm/car) 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0 Na rnet Su ia T n ts ra m Sw ana . (S nd Co C ) ng FM Gh p. Et ana . R B ia . (C in Ni K) ria E Ke n nz a Ca nia Si ail Za rail Tr bia Ug rail da Na ra M la ad i l M law ai C Ta ny FC da bo ib op ca Re ep en Bo aza i FM NC r ar ge Be an ep ila a m ta s m m w Ga an a oo hi o, . R az Sp m De De o, o, ng ng Co Co state-owned railways concessioned railways Source: Bullock 2009. Note: CFM = Caminhos de Ferro de Moçambique; FCE = Fianarantsoa Côte Est Railway. No data are given for Swaziland and FCE because they are essentially transit railways operated by other companies' equipment. Figure 3.11 Freight Wagon Productivity (Average, 1995­2005) 1.5 wagon productivity (million ntkm/wagon) 1.0 0.5 0 Na rnet Su ia T n ts ra m Sw ana . (S nd Co C ) ng FM Gh p. Et ana . R B ia . (C in Ni K) ria E Ke n nz a Ca nia Si ail Za rail Tr bia Ug rail da Na ra M la ad i l M law ai C Ta ny FC da bo ib op ca Re ep en Bo aza i FM NC r ar ge Be an ep ila a m ta s m m w Ga an a oo hi o, . R az Sp m De De o, o, ng ng Co Co state-owned railways concessioned railways Source: Bullock 2009. 121 Note: CFM = Caminhos de Ferro de Moçambique; FCE = Fianarantsoa Côte Est Railway. No data were available for Madagascar-FCE. Some railways were only concessioned close to or after 2005. This figure should therefore not be used to assess the effect of concessioning on rolling-stock productivity. 122 Africa's Transport Infrastructure spare parts. Wagon availability for a busy railway should therefore be well over 80 percent. The average annual distance that a wagon travels depends on its cycle time (the interval between successive loadings), which in turn is a function of the efficiency with which both railway operators and customers load and unload the wagons. In general, a wagon on a well-run railway should be able to travel at least 50,000 km per year.19 Given an average wagon load (excluding empty running) of about 20 tonnes on the region's 15-tonne-axle-load systems and about 30 tonnes on its 18 tonne systems, a well-run railway that is running 50,000 km per year per wagon should easily achieve annual productivity of 1 million ntkm per wagon. In prac- tice, the low productivity of many of the region's railways for 1995­2005 reflects very low annual wagon usages of 10,000 km or less, the result of excessively large (and generally obsolete) wagon fleets. As can be seen from figure 3.11, 11 railways out of 26 (around 42 percent) achieve less than 300,000 ntkm per year (Sudan, Democratic Republic of Congo­SNCC, Republic of Congo, Ghana, Ethiopia, Democratic Republic of Congo­CFMK, Nigeria, Kenya, Zambia, Uganda, Malawi). Impact of Concessioning on Productivity Concessioned railway companies in Africa have higher labor and asset productivity than state-controlled companies. In fact, the labor productiv- ity of concessionaires is on average twice that of their state-owned coun- terparts (Phipps 2008). Changes in labor and asset productivity for four railways over the five years prior to concessioning and in the period since are illustrated in figure 3.12, using four key indicators: (i) labor productivity (traffic units per employee), (ii) locomotive productivity (traffic units per locomotive),20 (iii) wagon productivity (net tonne-km per wagon), and (iv) carriage productivity (passenger-km per carriage). Labor productivity increased in each concession. Camrail's labor productivity rose sharply upon concessioning as traffic grew, but it then stabilized after about three years. It now appears to be increasing once again. The civil war in Côte d'Ivoire caused service suspensions for Sitarail in 2003­04, which interrupted the operator's upward trend until the situ- ation stabilized. After concessioning, CEAR retained only two-thirds of its staff, and at the same time, traffic grew by about 30 percent on an adjusted annual basis. Its productivity rose sharply as a result. The col- lapse of the Rivi Rivi Bridge left the northern half of CEAR's network and its associated staff with very little traffic, and consequently the operator's Figure 3.12 Labor and Asset Productivity for Four Railways before and after Concessioning a. Labor productivity b. Locomotive productivity 0.7 60 labor productivity (million TU/staff member) loco productivity (million TU/locomotive) 0.6 50 date of 0.5 concessioning 40 0.4 30 0.3 Rivi Rivi Bridge collapse 20 0.2 0.1 10 civil war in Côte d'Ivoire from 0 late 2003 0 ­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5 6 7 8 9 ­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5 6 7 8 9 years since concessioning years since concessioning c. Carriage productivity d. Wagon productivity 7.0 1,200 wagon productivity (000 ntkm/lwagon) Sitarail outsources carriage productivity (million pkm/car) 6.0 passenger services 1,000 5.0 800 4.0 600 3.0 400 2.0 1.0 200 0 0 ­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5 6 7 8 9 ­5 ­4 ­3 ­2 ­1 0 1 2 3 4 5 6 7 8 9 years since concessioning years since concessioning Sitarail Camrail CEAR RSZ 123 Source: Bullock 2009. Note: Year of concessioning = year 0; index at year 0 = 100. Black vertical line represents year of concessioning. RSZ = Railway Systems of Zambia. 124 Africa's Transport Infrastructure labor productivity fell in 2003­4.21 With CEAR's takeover of the Nacala line in 2008 and the reopening of the Rivi Rivi Bridge, these figures should improve. Most of the other recent concessions are likely to expe- rience similar increases in labor productivity, based on staff cuts: Madarail reduced its workforce by 50 percent, both the Mozambique concessions by about 60 percent, the Zambian concession by over 30 percent, and the Senegal concession by about 40 percent. Sitarail and Zambia experienced sharp increases in locomotive produc- tivity after concessioning, mostly due to the scrapping of their surplus equipment. Traffic growth in Cameroon resulted in steadily increasing productivity. The introduction of more powerful locomotives in 2007 approximately halved Camrail's fleet size, and productivity therefore doubled (not shown in figure 3.12). Locomotive productivity for CEAR has exhibited little change; in reality, however, the operator has not used much of its fleet and has bought some locomotives solely as sources of spare parts. Carriage and (especially) wagon fleets also showed improve- ments in productivity following concessioning, a result of traffic growth and scrapping of surplus stock. The productivity increases stemming from staff cuts and the scrap- ping of rolling stock after concessioning make the managers of the former government-run railways look less effective than they actually were. In many cases, key managers remained after concessioning; the government railways' large surpluses of labor often reflected not managerial incom- petence but political decisions to protect employees irrespective of rail- way efficiency. Concessionaires, almost without exception, also operate their railways with better asset utilization. In some cases, this improvement reflects greater use of assets that were previously lying idle (this is particularly the case with wagons). In others, it reflects the fact that surplus assets that could not be written off by the state companies for bureaucratic reasons were not taken over by the concessionaire. Service: Slow, Unreliable, and Unsafe The key determinants of service quality for both passenger and freight serv- ices are adequate capacity and frequency, safety, security, cleanliness, speed, and reliability. Based on these criteria, the service quality of many railways in Africa is poor. While concessioned railways cannot necessarily reduce transit times, they generally try to improve other aspects of service quality. Safety in particular is a cause for concern across the region. In the past 10 years, several major accidents have occurred, resulting in significant Railways: Not Pulling Their Weight 125 casualties. Some of these accidents were due to basic operating failures. Derailments, while a potential hazard on all railways, are extremely fre- quent in Africa. Although many occur at slow speeds in rail yards and pose a minimal safety threat, most African railways report over 100 derail- ments per year and some report 200, even 300. By comparison, the whole of the U.S. railway system typically has around 2,000 derailments each year, of which one-third occur on main lines, not in rail yards or on sid- ings; Canada has around 150 mainline derailments each year, and India less than 100. Relative to traffic volume, derailment rates in Africa are at an order of magnitude greater than most other regions, even allowing for track quality: U.S. Class 2 track (which has a 40 km/hr speed limit and is generally equivalent to the poorer segments of the mainline network in Africa) has one derailment for every 10 million wagon-km, while the rate for railways in Africa is some 30 times higher. Financial Performance Although almost all state-owned railways in Africa produce annual accounts, these are generally of little value other than as official records of revenue earned and expenditure made. Cash shortfalls are generally made good by grants from governments (often included as revenue). Most rail- ways just about break even on a cash basis, after receipt of government support, but there is almost always a substantial deferral of necessary maintenance. Depreciation may be recorded on a historic or replacement cost basis, but, as a noncash item, it is of little practical consequence for most railways. When the maintenance backlog becomes too great, it is typ- ically addressed using a loan from the government, with the expenditure listed in the books as an investment. The profitability of commercial concessionaires can be assessed a little more realistically.22 Sitarail and Camrail, which have been conces- sioned the longest, both make modest profits (Pozzo di Borgo 2006). CEAR in Malawi suffered long delays in finalizing the companion CDN concession in Mozambique and operated with working losses for several years around 2001. The performance of the Railway Systems of Zambia (RSZ) is unknown, and the Kenyan and Tanzanian concessions have not been in existence long enough to draw legitimate conclusions. Because of the lack of reliable information on the profits and losses of these concessionaires, a disaggregated approach to evaluating their financial performance, looking separately at the major elements of cost and rev- enue, is appropriate. 126 Africa's Transport Infrastructure Costs per Traffic Unit A railway's transportation costs per unit of traffic are determined by several factors: the unit costs of functional activities (track maintenance, train operations, and so on), traffic type, the efficiency of resource uti- lization, and the load factors achieved. The average cost per traffic unit across several railways in Africa between 2000 and 2005 is illustrated in figure 3.13. Unit costs for most railways lie in the range of 2­5 cents per TU. Costs are lowest for Botswana--a relatively flat railway with modern equip- ment, good track conditions, and a base mineral traffic. They are highest for OCBN and Tazara. Disaggregated Real Profitability Analysis The following sections provide a more detailed analysis of costs, revenues, and overall financial performance for three railways--two from low- income countries (Tanzania and Zambia) and one from a middle-income Figure 3.13 Cost per Traffic Unit for Railways in Sub-Saharan Africa (Average, 2000­05) 0.08 0.07 0.06 US$ per traffic unit 0.05 0.04 0.03 0.02 0.01 0 a l BN il CO a ia ra a AR N FB M a ai ra an ny bi an CD an CF za r CF CE OC CF ta m m Ke Gh w Ta nz Si Ca Za ts Ta Bo railway Source: Bullock 2009. Note: CDN = Corredor de Desenvolvimento do Norte; CFM = Caminhos de Ferro de Moçambique.; OCBN = Organisation Commune Benin-Niger des Chemins de Fer et des Transports. Railways: Not Pulling Their Weight 127 country (Botswana). All data are from the period 2000­02, before the concessioning in Tanzania and Zambia. Costs for railways can be divided into four categories: 1. The avoidable costs of train operation. These are the costs that could be cut if some services ceased operation. They can conveniently be measured as the costs of train crew, rolling-stock maintenance, fuel (or power), and passenger handling (ticket selling/commission and some station staff).23 2. Rolling-stock capital renewal costs.24 These can be derived by con- verting the original capital cost into an equivalent annual sum based on asset life (which differs according to the characteristics of individ- ual assets) and, using a real discount rate, assuming a 4 percent real rate of return. These are effectively sunk costs.25 3. Infrastructure operation and maintenance costs. These are referred to as "access charges" in this report. There are many methods for determining what portion of these costs should be charged to each type of service; for simplicity, a straightforward full-cost allocation method is used here. 4. The capital cost of infrastructure renewal. The first two of these categories are together referred to as the "above- rail costs." Estimates of revenues and costs will be used to assess the finan- cial performance of the railways, disaggregated into passenger and freight services. Passenger Service Profitability None of the three railways studied can cover its above-rail costs for its passenger traffic. The only one to come close is Tanzania, which has the highest earnings per carriage-km; the other two cover only about 50 per- cent of their above-rail working expenses. Even Tanzania performs poorly when rolling-stock capital renewal is included in costs, based on current usage of rolling stock. When access charges are included (but not the cap- ital cost of infrastructure renewal), the railways recover only between 20 and 40 percent of their costs. These figures might improve to between 30 and 65 percent with improved utilization of track and rolling stock (by halving the depreciation per unit of traffic). Some railway tariffs (including those in Tanzania) are essentially admin- istered within a government regulatory framework that considers only a subset of total costs. Nevertheless, many of the more poorly performing 128 Africa's Transport Infrastructure railways in the region would be unable to cover the above-rail costs of their existing set of passenger services even if they had the freedom to set their own tariffs. The cost recovery for passenger services on the three rail systems in 2002 is shown in figure 3.14. In the early 2000s, long-distance passenger railways needed to earn around $1 per carriage-km to be financially viable in the long term. Earnings of $0.75 per carriage-km would have covered avoidable costs of operation and a reasonable amount of the periodic maintenance required for the rolling-stock capital renewal costs. Third-party grants would have been needed to fund renewal costs of rolling stock (such as new locomo- tives and new carriages). Most economy-class coaches in the region can carry about 80 passengers, and a dynamic load factor of 70 percent is a reasonable if somewhat ambitious assumption. Railways would thus require a minimum tariff for third-class travel of 1.5­2.0 c/pkm, below which they require government support.26 Freight Service Profitability In contrast to passenger services, railways normally earn enough to cover the avoidable operating costs of freight services and sometimes enough Figure 3.14 Passenger Service Cost Recovery, 2002 100 90 80 70 60 percent 50 40 30 20 10 0 Tanzania Zambia Botswana railway above rail (AR) AR + depreciation AR + depreciation + access Source: Bullock 2009. Railways: Not Pulling Their Weight 129 to cover rolling-stock capital costs and infrastructure cost. The cost recovery for freight services in 2002 is shown for the same three railways in figure 3.15. Each of the three railways earned enough in 2002 to easily cover their above-rail costs and most of their above-rail depre- ciation and access costs. Only Botswana had a cost recovery of over 100 percent, but even it did not earn enough to cover the cost of renewing infrastructure. In 2002, railways would need to earn $0.80­$1.00 per wagon-km on their freight services, with operating costs of $0.60­$0.80 to be fully self-sustaining.27 (See Bullock 2009 for the calculations supporting this conclusion.) Concession Financing: Toward a New Structure The majority of concession financing is provided by governments, which in turn get such financing via low-interest sovereign loans from IFIs, usu- ally on terms that are not commercially available.28 Concessionaires pro- vide a relatively low percentage of equity (see figure 3.16). In most cases, the value of rolling stock transferred to the concession- aire outweighs the small amount of equity contributed by concession- aires (despite the poor condition of the rolling stock). As a result, the Figure 3.15 Freight Cost Recovery, 2002 percentage of different cost categories recovered 250 200 150 100 50 0 Tanzania Zambia Botswana railway above rail (AR) AR + depreciation AR + depreciation + access Source: Bullock 2009. 130 Africa's Transport Infrastructure Figure 3.16 Financing Structure of Select Concessions 180 100 160 90 percentage of assets funded by debt 140 80 70 120 60 US$ million 100 50 80 40 60 30 40 20 20 10 0 0 Madarail Sitarail Beira Camrail Transrail KRC URC Zambia TRC equity IFI/bilateral other % debt Source: Bullock 2009. Note: The Madarail concession reflects the September 2005 restructuring of an IFI loan for 21 million on-lent by the government to Madarail via a grant, which is included in the IFI/bilateral funds listed in the figure. Since the Sitarail concession is an affermage leasing the assets from a state holding company, Sitarail does not carry the debt contracted by the state holding companies ($56.7 million out of $63.6 million in total debt). Nevertheless, because Sitarail services the debt, it has been included in the figure. public sector assumes a significant portion of the financial risks associ- ated with infrastructure investment. This situation reflects the weak financial basis of many of the concessions, which are prone to significant liquidity problems and which cannot support major investment on a commercial basis. Major-asset maintenance and reinvestment are there- fore recurrent problems. Concession fees often have two components: variable (generally a per- centage of gross revenues) and fixed. In some cases, fees reflect the cost to government of providing assets to a concessionaire (as in a leasing agree- ment). More often, however, they are designed to ensure that private oper- ators share their revenues with the government. Successful railway concessions also normally pay taxes (such as a value-added tax, personnel social taxes, and income tax), the sum of which exceeds concession fees when taken over the projected lifespan of many concessions. Over their projected operational life spans (typically 20 to 25 years), concession fees and income tax each range from 2 to 14 percent of gross revenues, while net profit margins range from zero for Madarail to 25 percent for Zambia (figure 3.17).29 Railways: Not Pulling Their Weight 131 Figure 3.17 Concession Fees 30 25 percentage of net revenue 20 15 10 5 0 Zambia Transrail KRC Camrail Beira Sitarail Madarail TRC railway variable concession fee fixed concession fee taxes net profit Source: Bullock 2009. Bid projections must be interpreted with care. In the case of Zambia, the fixed component of the concession fee actually increased steadily throughout the concession period. The agreement stipulates that the con- cessionaire is responsible for paying the fee in full only if traffic levels are within 3 percent of the very ambitious traffic projections included in the reference financial model; otherwise, the fee will be adjusted downward. Based on these conditions, it is unlikely that the Zambian government has received any substantial payment from the concessionaire. Interestingly, the well-established Camrail and Sitarail concessions tend to project modest profit margins. These contrast starkly with the more optimistic forecasts of the newer concessions of Zambia, Kenya/Uganda, and Tanzania, whose projections of aggregate returns to government (through concession fees and taxes) and to concessionaires (in the form of profit) range from 35 to 50 percent of net revenue--equivalent to an operating ratio of 50 to 65 percent.30 Very few railways worldwide achieve similar ratios, including larger systems with modern equipment and much denser traffic than is carried by African railways. Governments should consider the combined impact of both taxes (pri- marily income tax) and concession fees when negotiating a concession, 132 Africa's Transport Infrastructure given the size of taxes and fees relative to the desired level of investment expenditure. A concession fee based on net revenue is undoubtedly easier to define,31 but income tax provides more flexibility, as the tax liability would automatically decrease in years in which there are unfore- seen difficulties. Regardless of the terms of the concession agreement, governments and advisers must realize that concessionaires can support only a limited amount in financial outflows--whether in the form of concession fees, borrowing costs, or rolling-stock acquisition costs. Proposed concessions that include high levels of both debt and concession fees are more likely to require renegotiation in the future. In the case of Madarail, for exam- ple, debt obligations outlined in the initial investment plan would have reached 18 c/ntkm in the fifth year after concessioning (2008), which would have been impossible to service given the railway's average rev- enue of only 5 c/ntkm. As a result, the government of Madagascar agreed to assume two-thirds of Madarail's debt in June 2005, after less than two years of operations. Similarly, Camrail's debt obligations would have reached about 8 c/ntkm in its fifth year. In 2005, Camrail and the govern- ment agreed on a concession amendment that transferred the cost of future track financing to the government until 2015 and capped the con- cession fee at 4 percent of net revenue. The Way Forward Concessioning of railways in Africa has generally improved their per- formance. But sustaining that improvement will depend critically on addressing a range of observed defects in the way concessions are designed and managed, and in the relationship between rail and road pricing policies. The Role of Concessions Railways that have not been concessioned--except for those immediately adjacent to South Africa (Botswana, Namibia, and Swaziland)--have con- tinued to deteriorate over the past decade. On the other hand, the asset and labor productivities of the concessioned railways have clearly improved. Better internal management has allowed concessionaires to streamline their cost and pricing structures, to seek new traffic, and to improve service qual- ity. With two significant exceptions (Zambia and Transrail), the railways have fulfilled the passenger service requirements of their concession agreements without raising the cost of passenger travel. They have also Railways: Not Pulling Their Weight 133 taken more realistic views about the future role of rail in the passenger markets, recognizing the limitations of rail in markets where it is unable to compete with faster road transport. Although many governments in Africa consider concessions a last resort, they still appear to be the way of the future. The Key Issues Few, if any, concessions have generated the cash flow needed to invest in infrastructure, and concessionaires are especially averse to investing in infrastructure with a life significantly beyond that of the concession. Probably most disappointing to governments has been the lack of infra- structure investment not funded by IFIs. Concessions in Africa are unlikely to be financially attractive to tradi- tional operators and instead appeal to companies that can secure financial benefits not directly linked to the railway operations (such as controlling an entire distribution chain or supplying of rail equipment). Generating increased interest from private operators will require changes in the mar- ket environment and financial structure of concessions, especially in five priority areas: Priority 1. Proper compensation arrangements for financially unviable passenger services Governments should provide operators with timely compensation for unprofitable passenger services. Any arrangement should be simple, easily auditable, and subject to periodic review. Priority 2. Improved capacity and willingness of private operators to finance track renewal These improvements could be achieved in two ways: · More realistic concession design and implementation, which would ensure that concessions (and therefore proposed track investments) were financially sound and that government's payment for the un- amortized value of the assets owed to the concessionaire at the end of the concession period was reasonable (providing that the concession agreement allowed for a possible extension of the concession period) · Independent finance of infrastructure renewal, perhaps through a land transport fund financed by both the road and rail sectors, into which concession payments could be made (instead of into the government general revenue fund) 134 Africa's Transport Infrastructure Priority 3. Effective and efficient regulation of private rail operators The main need here is to strengthen the regulatory bodies' capacity as well as to impose annual independent financial and operational audits as part of concession contracts. The regulatory bodies could be funded through concession fees or a land transport fund. Priority 4. A consistent and professional government approach to railway concessions Such an approach would require a properly staffed and funded oversight body with sufficient authority to control government action toward pri- vate rail operators. It should have ready access to experts in railway tech- nology and finance. Finally, the body should monitor the concession and report to the country's ministers of transport and finance. Priority 5. A consistent policy toward infrastructure use across modes As shown in chapter 2, road users are often charged fees too small to cover the costs of required maintenance. This allows road operators to charge artificially low rates, which limits the rates that railways can charge for freight transport. As a result, concessionaires have lower revenues and thus fewer funds to maintain and upgrade rail infrastructure. A consistent policy to address this problem should create a fairer and more competi- tive transport environment in Africa. Notes 1. The main source document for this chapter is Bullock (2009). Other impor- tant sources are Pozzo di Borgo (2006) and Mbangala (2001). 2. Rail gauge is the distance between the inner sides of the heads of the two par- allel rails that make up a single railway line. 3. The traffic units carried by a railway are defined as the sum of the passenger-km and the net tonne-km carried. This is a widely used standard measure, although it has some limitations as an indicator (for example, a first-class passenger-km in a French train à grande vitesse is treated the same as a passenger-km in a crowded suburban train). The relative weighting of passenger and freight is con- ventionally taken as 1:1, although alternative weightings have been used on some railways, usually in an attempt to reflect relative costs. 4. This situation also occurs at peak periods on much larger railways. For exam- ple, Spoornet endured heavy criticism in 2007 and 2008 for lacking sufficient capacity to carry coal and mineral exports to ports. 5. Years with abnormal events (such as wars and cyclones) were excluded from the averages. Hence the figure for Société Nationale de Chemins de Fer Railways: Not Pulling Their Weight 135 Congolais (SNCC) covers only a very short period, as the railway was not in operation from 1995 to June 2004. 6. The dynamic load factor is the ratio of total passenger-km to total seat-km. Not all passengers travel to the end of a route, so occupancy is by definition much higher at the maximum load point. 7. Tariffs (in terms of 2008 U.S. dollars) were greater than these averages in many cases; the 2008 freight yield for Camrail was 9.7 c/ntkm, for Sitarail 6.4 c/ntkm, and for Transrail 8.0 c/ntkm. These may not be sustainable as fuel prices fall but it seems they will remain above the 10-year average. 8. This structure also reflects the relative densities of these traffics: a wagonload of coal will generally weigh more and cost more overall to haul--but less per net tonne--than a wagonload of textiles. Railways therefore normally charge a comparatively high rate per tonne for low-density freight. 9. Much of the long-distance freight in southern Africa is carried on large, double- trailer, seven-axle combination rigs, which have a nominal maximum gross vehicle mass of 56 tonnes. Typical payloads for dense loads such as cement or steel are 30 to 40 tonnes. 10. If the railway does not pay the bond, the wagons are detained while a message is dispatched to Lusaka, the consignee deals with the bank, and the documen- tation is returned. It could take some weeks before the traffic is cleared, tying up wagons in the interim. 11. Concessions do not always include the entire network. In some cases, branch lines were excluded, such as the Mulobezi and Njanji branches in Zambia, the Lumbo branch near Nacala in Mozambique, and the St. Louis branch in Senegal. 12. An exception is the Sitarail affermage, where the assets were leased to the con- cessionaire and responsibility for investment remained with the government. 13. Some countries had an independent government inspector of railways. The inspectors were frequently promoted from the railway, however, and desired to return after their government tenure. Such an arrangement discouraged an honest inspection of incidents for which the inspector's future superior would be ultimately responsible. 14. Many contracts (for example, those for Société d'Exploitation Ferroviaire des Industries Chimiques du Sénégal [SEFICS] in Senegal and Magadi in Kenya) include clauses that allow third parties to operate on the concessioned infra- structure. The Camrail and Sitarail concession contracts include usage exclu- sivity periods of five and seven years, respectively, during which third parties cannot operate trains on their networks. Others, such as Madarail and Transrail, allow access from the start. 15. And even here there is the long-term threat of the alternate rail route through the Republic of Congo. 136 Africa's Transport Infrastructure 16. This excludes the special case of Sizarail in the Democratic Republic of Congo, which was terminated following a military coup. 17. A Chinese consortium that bid for the Beira concession, which was awarded to RITES in 2004, had a calculated return on equity of only 2 percent in its financial proposals. 18. Comparisons of railway productivity should be made with care: the number of staff that a railway employs is a function of how much work (especially major-asset maintenance) the railway outsources and how much it keeps in house. Also, the almost universally used measure of work done by railways is traffic units (the sum of net tonne-km and passenger-km), which can be an unreliable indicator, as servicing one passenger-km generally requires more resources than one net tonne-km. Measures of productivity based on traffic units therefore favor railways that primarily transport freight, and particularly those transporting a high proportion of minerals, which are heavy and require only simple servicing. A rate of 1 million TUs per employee would be good in African circumstances. Transnet Freight Rail is helped by its heavy-haul lines and also because it has dropped most of its passenger services. Very few devel- oped countries outside those having specialized long-haul freight railways have reached a rate of 3 million TUs per employee--most Western European countries are around 700,000 to 1 million TUs per employee, but this rate is affected by heavy passenger volumes. 19. It should be able to travel much more if the train carries a single commodity between only two locations. In the early 1980s, such a "block train" (a train run as a single unit and not at any time split up) ran on Zimbabwe's railways, carrying coal between Hwange and the Zisco steel plant at Kwekwe. The wag- ons on these trains traveled around 200,000 km per year. 20. The CEAR locomotive productivities ignore the scrap locomotives that were purchased for spare parts, which tend to inflate the productivity figure. 21. These statistics illustrate the difficulties with such broad measures of produc- tivity. Because of the traffic shortages caused by the bridge collapse in 2004, CEAR locomotives did 25 percent of their work on hire to Caminhos de Ferro de Moçambique (CFM), which is not reflected in the traffic statistics. 22. Some suggest that although concessionaires publish low profits, they have charged significant fees for providing management services and the like. See Pozzo di Borgo (2006). 23. A proportion of infrastructure costs also depends on usage and should, strictly speaking, be included. But for many of the more basic low-density railways, the incremental impact of passenger services on infrastructure is relatively small unless the services are suspended, in which case significant changes in track and signaling standards can often be made. 24. These costs have been termed "above-rail depreciation," assuming that depre- ciation is based on the renewal cost of assets rather than the historic cost. Railways: Not Pulling Their Weight 137 Costs calculated in this way are unlikely to appear in the accounts of railways in Africa. 25. Most rolling stock in use in Africa--except South Africa--is not in good enough condition to be resold. 26. This figure is obtained as follows: 80 seats per carriage at 70 percent dynamic load factor (that is, passenger-km: seat-km) is equal to 56 pkm per carriage- km (maximum). If the earnings needed were $1.00 per carriage-km (mini- mum) to include rolling-stock renewal, then the minimum tariff would be $1.00/56, or 1.8 c/pkm. This is rounded to 1.5­2.0 c/pkm. Lower load factors (which are a function of service frequency and train size) would increase the minimum, although in most cases it would still be comparable with bus fares. Service frequency and travel time remain the limiting factors for railways. 27. Excluding concession fees, typically around $0.05 per wagon-km. 28. Many other developing countries use the same practice to loan to publicly owned railways. In the mid-1990s, on-lending was usually made at a premium (for example, the International Development Association provided Côte d'Ivoire with a loan at 0.75 percent interest, which the government on-lent at 8 percent interest to Sitarail for its concession). To attract operators, subsequent concession loans featured sharply reduced premiums--as low as 0 percent in the case of Madarail. As a result, the average interest on the Madarail operator's debt is only 1.73 percent, with a 7-year grace period and a 25-year tenor. 29. Net profit margins equal total operating revenues minus total operating costs minus depreciation and interest on debt capital minus taxable income. 30. Operating ratio equals expenses divided by operating revenues. 31. Concessionaires can extract funds under the guise of costs in a number of ways, such as inflating "technical assistance" fees for providing management. References Bullock, R. 2009. "Railways in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 17, World Bank, Washington, DC. Mbangala, M. 2001. Le transport ferroviaire en Afrique noire: Fonctionnement, per- formances, perspectives. Liége, Belgium: Editions de l'Université de Liége. Mbangala, M., and S. Perelman. 1997. "L'efficacité technique des chemins de fer en Afrique Sub-Saharienne: une comparaison internationale par la méthode de DEA." Revue d'Economie de Développement 3 (93): 91­115. Murdoch, Jill. 2005. "Assessing the Impact of Privatization in Africa--Case Study of Camrail." World Bank, Washington, DC. Phipps, L. 2008. "Review of the Effectiveness of Rail Concessions in the SADC Region." United States Agency for International Development, Gaborone, Botswana. 138 Africa's Transport Infrastructure Pozzo di Borgo, P. 2006. "Review of Selected Railway Concessions in Sub-Saharan Africa." Africa Region Report, World Bank, Washington, DC. Thompson, L. S. 2007. "Spoornet and Transnet Sectoral Reference Paper." Thompson, Galenson, and Associates, Washington, DC. UNCTAD (United Nations Conference on Trade and Development). 2003. Efficient Transport and Trade Facilitation to Improve Participation by Developing Countries in International Trade. Geneva: UNCTAD. CHAPTER 4 Airports and Air Transport: Policies for Growth A viable, stable air transport industry is critical to Africa's integration into the global economy.1 In an increasingly liberalized world air-transport market, however, many of Africa's indigenous air transport operators-- state-owned flag carriers in particular--have failed. While the region's air traffic is growing, overall connectivity is not, and regional carriers have the worst safety record in the world. Furthermore, air traffic control (ATC) and airport infrastructure are inadequate. Fortunately, there are promising signs. New carriers are entering the intercontinental market and are beginning to realize the benefits of market liberalization. The challenge now is to build on these successes. Airport Infrastructure The discussion of air transport begins by looking at the capacity, condi- tion, and utilization of airport infrastructure as well as the status of ATC and navigation aids for commercial air traffic in Africa. Service and Connections: Falling Off Africa has at least 2,900 airports.2 Yet less than 10 percent of these receive scheduled services. Moreover, that number is falling. While 318 airports 139 140 Africa's Transport Infrastructure received scheduled services in 2001, only 280 airports were estimated to be receiving services in November 2007, and for only 261 of these was service throughout the year. With the exception of airports in the Banjul Accord Group (BAG), between 20 and 40 percent fewer airports received scheduled services in 2007 than in 2001 (see appendix 4b).3 The level of international connectivity, measured in terms of the num- ber of airports receiving direct international service, is also in decline (see map 4.1 and appendix 4c for shifts in overall connectivity). Air Traffic Control and Navigation: Inadequate and Poorly Financed Ground-based navigation installations are sparse in Africa (see appen- dix 4d). The main corridor in the east of the continent--stretching Map 4.1 International Connectivity in Sub-Saharan Africa: Winners and Losers 2001­07 Loss in connectivity Gain in connectivity Not part of sample Source: Analysis by H. Bofinger. Airports and Air Transport: Policies for Growth 141 from South Africa to the Arab Republic of Egypt--has the most ATC installations. Ghana, Nigeria, Tanzania, Uganda, and Zimbabwe have some. But the rest of Africa, including Ethiopia--one of the region's important hubs--lacks coverage. Malawi once had some installations, but the equipment became too expensive to maintain and fell into disrepair; it is no longer salvageable. Even the existing ATC installations do not necessarily use radar sep- aration, which is a technique for issuing directions and headings to air- crafts based on radar images. East Africa is typical. In Kenya, only Nairobi uses full-time radar vectoring,4 while Mombasa switches to radar procedures only if weather conditions demand it. Tanzania has a good radar installation in Dar es Salaam, with a secondary radar range in excess of 300 kilometers (km), but lacks radar-certified controllers and therefore cannot use radar vectoring. The Ugandan military pro- vided the country with radar services, using aged technology, until a new civilian system was installed in 2008. This unsatisfactory pattern is common throughout Africa. The safety risks that it entails will only grow as traffic increases. The lack of radar coverage in the region is not an insurmountable chal- lenge. Modern surveillance technology is moving away from radar instal- lations toward the more advanced (and cheaper) automatic dependent surveillance-broadcast (ADS-B). In this system, the aircraft determines its position using a global navigation satellite system and transmits it to a ground station, which then relays it to the ATC center. Positions obtained using modern global positioning system (GPS) technology can be accu- rate within 30 meters, avoiding the challenges of using radar technology to locate aircraft accurately at long distances and to detect changes in their speed. ADS-B helps separate aircraft, which is not a problem in lightly trafficked areas, and provides important navigation information to pilots. Some ADS-B systems also allow nearby aircraft to broadcast their positions to one another, provided they have the proper equipment. ADS-B is being considered in a planned redesign of ATC in the Southern African Development Community (SADC) region. The air transport sector could clearly benefit from this new surveillance tech- nology, which is only about a quarter of the cost of radar systems and has lower maintenance costs. In future, most aircraft will probably have their own GPS, and airports will learn to take advantage of the technol- ogy as it becomes more widespread. Most airports in the region with an estimated capacity of 1 million passengers or more have an instrumented landing system (ILS).5 They 142 Africa's Transport Infrastructure are far less frequent in smaller, older airports, however, where outdated nondirectional beacon systems are still prevalent. Today, satellite technol- ogy provides low-cost replacement options for many ground-based navi- gation systems. Nevertheless, it appears that in many cases either no plans have been made or no funding has been obtained by national civil avia- tion authorities (CAAs) for the replacement of increasingly obsolete technologies (Schlumberger 2007). Airport Infrastructure Capacity The effective overall capacity of an airport is determined by the facilities with the lowest capacity; such facilities may be airside (such as runways and airport parking space) or landside (such as terminals). Airside capacity: Capable of economical expansion. A single-runway air- port operating with a five-minute lag between flights could theoretically accommodate 144 flights in 12 hours--equivalent to over 1,000 flights a week. With an average passenger load of 120, such an airport could serv- ice over 17,000 passengers a day, or over 6 million passengers per year. Yet Johannesburg International Airport is the only airport in the region whose traffic volume exceeds this figure. Even assuming a 20-minute lag between flights, the theoretical capacity of a single-runway airport would be over 1.5 million passengers per year. Based on a comparison of theo- retical runway capacity and actual air traffic (appendix 4e), Africa already has sufficient potential airport capacity. Existing airports should therefore focus on maximizing their effective capacity. When extra capacity is needed, rehabilitation is generally more economical than new construc- tion (see appendix 4f). Many African airports have a low-cost design that limits runway capacity. On landing, aircraft must use a turning bay at the end of the runway and taxi back up the main runway to the airport ramp, or apron, where they are parked, loaded and unloaded, refueled, and boarded. The access to the apron is usually in the center of the run- way. This arrangement is fine if there is enough time between depart- ing and arriving aircraft to complete the procedure, but high-volume airports require parallel taxiways with multiple turnoff ramps from the runway. A common, and economical, solution for airports with the turning bay configuration is to construct a parallel taxiway onto which aircraft move at the end of the runway. This makes a five-minute lag between flights quite manageable. An additional constraint is that airports Airports and Air Transport: Policies for Growth 143 often have narrow peak periods of aircraft arrivals, which puts pres- sure on both runways and terminals. In this case, flights should be rescheduled to avoid unmanageable traffic. Minor investment in taxi- ways and better management of capacity can thus obviate the need for major investments in new runways in most African airports in the near future. Landside capacity: Careful management required. Inadequate passen- ger terminal capacity is more common. Data on passenger throughput of African airports are surprisingly sparse. Table 4.1 therefore gives firm figures only for years for which the data is thought to be most reliable, which explains the large number of empty cells in the table. A comparison of reported terminal capacity and passengers at African airports shows that several are operating at or above design capacity (table 4.1). Some airports have already begun to address the problem. For example, Nairobi's airport is upgrading its terminal to accommo- date over 9 million passengers. Decisions to upgrade airport terminals, however, must be made on a case-by-case basis. Airport planners usually assume the need for 20 square meters (m2) of terminal per international traveler at any one time, or between 0.007 and 0.010 m2 multiplied by the total number of annual passengers. But such formulae must be treated with caution, as space requirements depend on airport use patterns. In particular, the pres- sure on terminal capacity will be a function not only of the maximum number of flights per hour but also of the ratio of that maximum to the weekly average. For the three airports with the highest number of maximum flights per hour (Johannesburg, Nairobi, and Lagos), the ratio of maximum flights per hour to average flights per week was less than three to one, while many of the smaller airports have ratios of more than six to one (see appendix 4b). In such circumstances, rescheduling arrivals and departures may be a much more economical solution to overcrowding of smaller airports than investment in extra terminal capacity. Infrastructure Conditions: Fraying at the Edges Nearly all of the airports with services in November 2007 had at least one paved major runway. Only 12 of these airstrips were unpaved, most in countries with recent or ongoing military conflict. It appears, however, that 25 percent of the 173 African airports for which satellite images 144 Table 4.1 Airport Terminal Capacity vs. Reported Passengers and Estimated Seats Reported passengers (millions) Reported 2007 estimated Country City Airport capacity (millions) 2000 2003 2004 2005 2006 2007 seats (millions) South Africa Johannesburg JNB 11.9 -- -- -- -- -- 19 25.3 Kenya Nairobi NBO 2.5 -- -- -- 4.3 -- -- 6.3 Mauritius Mauritius MRU 1.5 -- -- -- -- 2.2 -- 3.0 Senegal Dakar DKR 1.0 -- -- -- -- -- -- 2.5 Tanzania Dar es Salaam DAR 1.5 -- -- -- -- -- -- 1.9 Zambia Lusaka LUN 0.4 -- -- -- -- 0.6 -- 1.3 Kenya Mombasa MBA 0.9 -- -- -- -- 1.0 -- 1.1 Zimbabwe Harare HRE 0.5 -- -- -- -- -- -- 1.1 Mali Bamako BKO 0.4 -- -- -- -- 0.5 -- 0.7 Djibouti Djibouti JIB 0.5 -- -- -- 0.1 -- -- 0.6 Rwanda Kigali KGL 4.4 0.1 -- -- -- -- -- 0.5 Nigeria Kano KAN 0.5 0.3 -- -- -- -- -- 0.4 Malawi Lilongwe LLW 0.2 -- -- 0.2 -- -- -- 0.4 Source: Bofinger 2009. Note: -- = not available. Airports and Air Transport: Policies for Growth 145 were available are in marginal or poor condition, and 21 percent are in poor condition. On the other hand, only about 4 percent of the region's air traffic passes through marginal or poor airports (see table 4.2). There is limited reliable information on the quality of airside infra- structure. For example, the ILS at Maseru International Airport in Lesotho is so unreliable that the scheduling integrity of the only air- line servicing the airport, South African Airlink Express, has been compromised.6 In other cases, modern global navigation satellite sys- tems may have been designed and financed, and are known by the financing agency to be complete and operational, but are not in the inven- tory of airside services and installations in the Aeronautical Information Manual or in databases such as that of Jeppesen.7 Airside infrastruc- ture is an area where more systematic collection and publication of data, either by the International Civil Aviation Organization (ICAO) or by the regional economic communities, could improve sys- tem performance. Airports with higher traffic volumes generally have higher-quality air- side infrastructure. Main hubs, such as Johannesburg and Nairobi, have adequate airside infrastructure, including standard runway length and ILS. Lower-quality infrastructure is much more prevalent among airports with low traffic volumes (below a million seats per year). For example, although relatively few airports in the region have unpaved runways, some countries have a high number of airports with poor runway conditions. Data from the International Air Transport Association (IATA) suggest that among African countries that have not recently undergone conflict, Tanzania stands out in having five airports using unpaved runways for advertised, regularly scheduled services. However, an informal domestic aviation sector in Africa, not recorded in the IATA data, may utilize even more unpaved runways. Table 4.2 Overall Runway Quality in Africa Rating Airports Percent Seats Percent Excellent 31 17 69,666,792 63 Very good 51 28 26,574,283 24 Fair 52 29 9,285,100 8 Marginal 8 4 2,291,844 2 Poor 37 21 2,419,054 2 Total 179 100 110,237,072 100 Source: Author's analysis of data collected by the World Bank. 146 Africa's Transport Infrastructure Airport Charges: High Airport landing charges are high in Africa, although they vary consid- erably by airport and by aircraft. Comparison of charges for three air- craft types across 15 airports in Sub-Saharan Africa and 3 in North Africa with FraPort in Frankfurt, Germany (figure 4.1), shows that charges at Sub-Saharan African airports were on average 30 to 40 percent higher than at FraPort, while those in North Africa were comparable or lower. (Charges were particularly high in Cameroon, Côte d'Ivoire, and Ghana, but even adjusting for those outliers, they averaged 29 percent higher at African airports.) That the discrepancy increases dramatically with aircraft size suggests that airports charge intercontinental travelers more, perhaps to generate foreign currency revenues. In some cases, passenger fees exceed $80 per passenger. On the other hand, airports in Sub-Saharan Africa rarely have other sources of revenue, such as shopping, car rentals, and duty-free con- cessions, which contribute a large proportion of overall airport revenues in industrialized countries. The higher charges in the region are therefore unsurprising. Figure 4.1 Airport Charges Overall, by Aircraft Type across 19 Sample Airports 30,000 25,000 20,000 landing charge 15,000 10,000 5,000 0 ria o Ca nin on ire n a ea a ria i ia co Ni r S e ia ra al e sia t M Mal ge or an ny on as bo an r Si neg in oc ge be ge lvo Cô ero ni aP Be aF Ni Ke Gh Le Ga rit Gu Tu or Al Li Fr d' m in au M te rk er Bu airport location A320 B767 B747 Source: ADPI 2008. Airports and Air Transport: Policies for Growth 147 Ownership and Management: Scope for More Private Sector Participation There are four ownership and operation schemes for airports: (i) pub- lic ownership and operation, (ii) regional ownership and operation, (iii) public ownership and private operation, and (iv) private ownership and operation. The first model is still common in Africa. Like governments of poor countries in other regions, the governments of countries in Africa often consider airports to be public infrastructure that provides revenue and foreign currency. Even if the airport is corporatized (as in the case of the Airports Company South Africa), the state retains majority ownership. However, even airports with operational surpluses fail to undertake nec- essary maintenance and reinvestment (Button 2008). The success of this model is thus questionable. Regional ownership is most common for secondary airports and is often used by central governments to remove less profitable airports from the national budget. It is also found in federal countries, especially those with strong airport markets, such as the United States. This form of ownership is unlikely to benefit regional airports in poorer countries with weak provin- cial government systems and is therefore generally inappropriate for Africa. There are several models for public-private partnerships (PPPs): joint ventures, partial and majority divestitures, management contracts, and concession contracts. Table 4.3 shows the few recorded attempts at PPPs in African airports. They have occurred in markets of all sizes: Cameroon, a small market that serves below 1 million seats a year; Tanzania, which serves more than 1 million seats a year (an average size for the region); and South Africa, the largest market in Sub-Saharan Africa. Cameroon was the only country in the region with a management contract covering a system of seven airports owned by the Aéroports du Cameroun. Its major stakeholders were Aéroports de Paris, with 34 percent, and the gov- ernment of Cameroon, with 24 percent (the remaining stakes were shared by other carriers and a bank). But when Aéroports de Paris failed to meet the agreed-upon requirements for funding the Douala International Airport rehabilitation in 2004, the government took over its share. This model has thus also proved less than perfect. The majority of private participation in African airports is through concessions. Under a concession agreement, the government continues to fund infrastructure investment while the concessionaire assumes respon- sibility for service provision. This arrangement allows private firms to 148 Table 4.3 Public-Private Investments in African Airports Financial Number Government closure Project Contract Terminal Multiple of granting Investment % Country year Project name Type of PPP status Location period year systems transactions contract year private Cameroon 1993 Aéroports du Concession Operational 7 airports 15 2008 Yes 7 Federal 1993 71 Cameroon Côte d'Ivoire 1996 Abidjan Concession Operational Abidjan 15 2011 No 1 Federal 1996 100 International Airport Djibouti 2002 Djibouti Concession Operational Djibouti n.a. n.a. No 1 2002 0 International Airport Kenya 1998 Jomo Kenyatta Greenfield Operational Nairobi -- -- No n.a. Federal 1998 100 Airport project Cargo Terminal Madagascar 1991 Aéroports de Concession Concluded 12 airports 15 2006 Yes 12 Federal 1991 34 Madagascar Mauritius 1999 Mauritius Airport Management Concluded Port Louie 5 2004 No 1 Federal 1999 100 and lease contract Nigeria 2006 Murtala Greenfield Construction Lagos 25 2027 No 1 Federal 2006 100 Muhammed project Terminal One South Africa 1998 Airports Divestiture Canceled Johannesburg, n.a. 2005 Yes 11 Federal 1998 20 Company Ltd. 11 airports South Africa 2000 Kruger Park Divestiture Operational Phalaborwa n.a. n.a. No 1 Federal 2000 100 Gateway Airport South Africa 2000 Rand Airport Divestiture Operational Gauteng n.a. n.a. No 1 Federal 2000 80 South Africa 2001 Mpumalanga Greenfield Operational Nelspruit -- -- No 1 State/ 2001 90 Airport project provincial Tanzania 1998 Kilimanjaro Concession Operational Kilimanjaro 25 2023 No 1 Federal 1998 100 International Airport Source: Analysis by H. Bofinger of data from World Bank 2009. Note: n.a. = not applicable; -- = not available. There are more PPPs, such as the partial privatization of the airport holding company Aéroport de Libreville, Gabon, in 1996. But most others are more concentrated on management contracts and specified services, rather than full operations of and investments in airports. 149 150 Africa's Transport Infrastructure offer specific services, such as SwissPort's passenger counter services in Johannesburg and Dar es Salaam and private contractors' cargo-handling functions in lesser-known airports, such as Mwanza in Tanzania. Contract bidding occurs in regular cycles, and terms vary from airport to airport. Concessioning of specific services is a well-developed model in airports throughout the world, and may be the most appropriate and sustainable form of PPP for Africa. Full privatization is rare among airports and is generally attractive only for airports with substantial passenger traffic that is potentially profitable. One example is that of the British Airports Authority, which owns the three main London airports (Heathrow, Gatwick, and Stansted) as well as the three main Scottish airports (Edinburgh, Glasgow, and Aberdeen). However, it has been argued that because of the authority's monopolis- tic nature, airport charges have soared, service quality declined, and rein- vestment in basic airport infrastructure has been insufficient (Osborne 2007). Following an inquiry by the UK Competition Commission, the authority was required to sell off Gatwick, Stansted, and either Glasgow or Edinburgh. The Airports Company South Africa, which owns 10 airports in South Africa, appears to follow this model. In practice, however, the company is controlled by the government of South Africa. Meanwhile, the privatiza- tion of complete airports is slowing worldwide: noticeably fewer transac- tions occurred in 2007 than in the immediately preceding years (ACI 2008). This trend is likely to continue. Regulatory Institutions: Struggling to Maintain Adequate Staff and Funding In most countries worldwide, a general aviation law establishes and authorizes regulatory bodies, which then implement necessary regula- tions. Many countries use U.S. Federal Aviation Administration standards. Air transport regulatory bodies generally comprise both a civil aviation authority and an airport operations organization (or organizations). In most cases, the CAA is either an agency of the ministry of transport (as in South Africa) or a statutory body under the sponsorship of the min- istry. The CAA typically provides ATC and navigation services and is responsible for safety oversight and certification of airports, aircraft, and personnel. In some countries, the CAA acts as a government adviser or government agent in international air service regulation. The airport organization typically provides or regulates all airport services, includ- ing instrumented landing facilities. Airports and Air Transport: Policies for Growth 151 CAAs are often set up as quasi-commercial bodies and publish annual reports and financial accounts. For example, the South African Civil Aviation Authority is funded by a combination of direct and indirect fees and direct government funding of its accident investigation functions. It charges direct fees to airports for its services, which are passed on to air- lines as a charge per passenger landed. It also levies a general aviation fuel tax. In some countries with large land areas located under major air routes, overflight fees charged to transiting airlines provide significant air navigation revenues and thus an important source of CAA funding. The allocation of such funding can be politically contentious. The revenues from services provided by a truly independent regulatory body would be received directly by that body and applied to the sector. In many cases, however, the government treasury receives the revenues, and the agency is forced to negotiate for its share. Two factors limit the effectiveness of regulatory bodies in safety over- sight. First, airlines can offer highly trained professional safety inspectors substantially higher salaries than the typical CAA in Africa, which cannot afford sufficiently capable staff. Second, regulation is subject to abuse by political influence. For example, a politically well-connected person may be allowed to operate an aircraft that does not meet safety standards and would not be allowed to fly in another country. Political autonomy and independent funding are therefore essential for effective regulatory bod- ies. The poor safety record of air transport in Africa can be attributed to a lack of both. Regional Safety Oversight Bodies: Filling the Gaps To address Africa's shortcomings in oversight, regions have begun to pool resources. With support from the U.S. Department of Transportation's Safe Skies for Africa Program,8 a regional CAA was recently formed in East Africa to augment the capacity of national air safety systems with pooled funds and shared staff. The Agence pour la Sécurité de la Navigation Aérienne en Afrique et à Madagascar (ASECNA) pools air navigation services and other infrastructure and manages eight airports in its 15 member countries. Based in Dakar, ASECNA manages 16.1 square km of airspace (1.5 times the area of Europe), providing services that are the responsibility of a CAA in many other countries. Finally, projects under the Cooperative Development of Operational Safety and Continuing Airworthiness Programme (COSCAP) of the International Civil Aviation Organization are being planned for several of the regional organizations in Africa, including SADC, the Economic and Monetary 152 Africa's Transport Infrastructure Community of Central Africa (CEMAC), the Common Market for Eastern and Southern Africa (COMESA), and BAG. The aim of these projects is to resolve regional safety oversight issues and to harmonize regulations on a regional basis. Operations Starting from relatively low levels by world standards, air transport service and patronage have been growing steadily in all sectors and in most regions. Intercontinental traffic is concentrated in three hubs. Liberalization of the international markets has increased market concentration, and there are an increasing number of low-density markets with a sole supplier. Protection of national flag carriers is a continuing impediment to growth in some countries. Equipment is getting younger and is adapting better to specific market demands. Traffic Rates: Low but Growing African air transport has experienced significant growth in the past decade, especially between 2001 and 2004 (table 4.4). Nevertheless, with just under 12 percent of the world's population, Africa still accounted for less than 3.7 percent of the global market in 2007. Market supply, which in 2007 consisted of roughly 72.3 million pas- senger seats, grew by an annual average of 6.2 percent between 2001 Table 4.4 Estimated Seats and Growth Rates in African Air Transport Markets Estimated Estimated Estimated seats, seats, seats, Growth, Growth, Growth, 2001 2004 2007 2001­04 2004­07 2001­07 Market (millions) (millions) (millions) (%) (%) (%) All Sub-Saharan Africa 50.4 54.5 72.3 2.7 9.9 6.2 Sub-Saharan domestic 18.2 19.4 27.5 2.1 12.4 7.1 Sub-Saharan international 11.8 11.9 14.3 0.3 6.5 3.4 Sub-Saharan intercontinental 19.5 22.1 28.1 4.1 8.4 6.2 Between North Africa and Sub-Saharan Africa 0.9 1.3 2.5 11.1 24.8 17.8 Source: Analysis by H. Bofinger of data provided by ADG Seabury. Airports and Air Transport: Policies for Growth 153 and 2007. Growth was lower between 2001 and 2004 but surged to 9.9 percent between 2004 and 2007. Although all types of traffic have experienced growth, intercontinental traffic, international traffic from certain hubs (Addis Ababa, Nairobi), and domestic traffic in certain countries, such as Nigeria, have grown most quickly. One of the least- connected regions lies between the countries in West and Central Africa and the better-developed network in the east. As liberalization spreads throughout Africa, however, major carriers from the east have begun to fill this gap. Growth in passenger traffic, measured in revenue passenger-km (rpkm), has mirrored supply growth. It grew steadily between 1997 and 2001 but experienced a mild downturn after September 11, 2001, followed by two more years of growth. The collapse of several African airlines in 2004 resulted in a significant reduction in intra- African traffic. New supply capacity entered the marketplace between 2005 and 2006, however, and traffic surpassed what was seen at the beginning of 2004. Passenger growth, like supply growth, has been highly uneven. East Africa and southern Africa have benefited from the growth and devel- opment of three key players: South African Airways (SAA), Ethiopian Airlines, and Kenya Airways. In contrast, passenger traffic in West and Central Africa declined significantly after the collapse of supply by several regional airlines (including Air Afrique) and has not yet fully recovered. (More information on capacity growth is contained in appendix 4e.) Intercontinental traffic: Heavy reliance on three regional hubs. Inter- continental capacity in Africa grew by 43.6 percent between 2001 and 2007, at an annual growth rate of 6.2 percent. In total, 158 carriers pro- vided intercontinental services in 2007, with an average of 3.45 airlines competing in each of the top 20 markets (map 4.2). Between 2001 and 2007, 50 operators left the market, of which Air Afrique, Swissair, and Ghana Airways had the most capacity. At the same time, over 80 operators entered--with nearly double the capacity of those that left. Service between South Africa and Egypt and the United Arab Emirates had the highest growth rates in terms of capacity offered along major routes. The only routes on which the number of intercontinental operators declined between 2001 and 2004 were between the United States and South Africa. The top five airlines--SAA, Air France, British Airways, EgyptAir, and Emirates--hold over 30 percent 154 Africa's Transport Infrastructure Map 4.2 Top 30 Intercontinental Routes for Sub-Saharan Africa as of November 2007 United Kingdom Netherlands France Germany Spain Sudan Morocco Saudi Arabia United States United Arab Emirates Cape Verde China Ethiopia Senegal Singapore Nigeria Kenya Cameroon Mautitius Tanzania Angola Madagascar South Africa Australia Source: Analysis by H. Bofinger. Note: Johannesburg serves as the most important entry point, with the three largest partners (excluding North Africa) being the United Kingdom, Germany, and the United Arab Emirates (UAE). of market share of all African intercontinental traffic. The top 20 airlines include 8 African carriers. Most intercontinental traffic in the region passes through one of three major hubs: Johannesburg, Nairobi, and Addis Ababa. The route between the United Kingdom and Johannesburg is the most heavily traveled. Between 2001 and 2007, service from the Middle East to all three African hubs increased significantly. In 2001, for example, the United Arab Emirates had only 2 of the region's top 30 country pairs; by 2007 it had 5. Traffic from the Johannesburg hub to East Asia and the Pacific regions nearly doubled between 2004 and 2007 to 1.6 million seats. International travel within Africa: Fewer connections, more passengers. International capacity within Africa grew by 3 percent per year between 2001 and 2004, and by 9 percent per year between 2004 and 2007 (figure 4.2). The acceleration from 2004 onward is accounted for by the entry of new carriers, as the industry recovered from the collapse of several African airlines that had caused a significant reduc- tion in intra-African traffic in 2004. Capacity growth was highest between Sub-Saharan Africa and North Africa, at 25 percent per year. International travel within Sub-Saharan Airports and Air Transport: Policies for Growth 155 Figure 4.2 Estimated International Passenger Capacity between 2001 and 2007, as Measured in Seat-Kilometers 35 30 25 capacity (billions) 20 15 10 5 0 2001 2004 2007 estimated seat-km SSA to SSA SSA to NA NA to NA Source: Analysis by H. Bofinger. Note: SSA = Sub-Saharan Africa; NA = North Africa. Africa, which accounts for the bulk of intra-African international travel, grew at 6.5 percent. Over the same period, however, connectivity within Sub-Saharan Africa declined as the number of country pairs with connect- ing routes fell from 218 to 190. The loss of connectivity can be attributed largely to the collapse of airlines, including Air Afrique (see box 4.1), Nigeria Airways, Air Gabon, and the Ghana Airways Corporation. This type of collapse is somewhat anomalous, as it occurred during a period of overall capacity growth. Overall, 31 airlines left the market between 2001 and 2007, taking with them a combined capacity of nearly 8 million seats. At the same time, 34 new operators entered the market, bringing nearly double the lost capacity (15 million seats). Connectivity is more developed in the eastern part of Africa, anchored by the major hubs in Johannesburg (South Africa), Nairobi (Kenya), and Addis Ababa (Ethiopia). These airports serve 36 percent of all interna- tional traffic within Africa (see table 4.5). Each has a dominant national airline: SAA, Kenya Airways, and Ethiopian Airlines account for 33 per- cent, 70 percent, and 83 percent of international traffic, respectively, at their hubs. The highest growth in intra-African travel was exhibited by the coun- tries of the BAG, including Nigeria, followed by the more developed regions of East and southern Africa, and North Africa. In contrast, because 156 Africa's Transport Infrastructure Box 4.1 Air Afrique Air Afrique was formed in 1961 as an African carrier headquartered in Abidjan, Côte d'Ivoire, and was owned by 12 West African countries, Air France, the Union Aéromaritime de Transport, and the Société pour le Développement du Transport Aérien en Afrique. The airline went from piston-engined propeller operations to jet-engined wide-bodies such as the Airbus 310 in the 1980s. Just as with flag carriers, the airline became a regional symbol of pride and inde- pendence. But quality of service was sometimes compromised even in the best of times, for example, when reservation systems collapsed, making seat assignments impossible. In its last days, passengers were increasingly stranded. Some claimed that prioritized seating was given to nonpaying passengers of importance and that scheduling integrity had diminished. Efforts by the airline's president to restructure the airline in 2001 by cutting jobs were vehemently opposed by its employees, who at one point refused to fly an airplane with the president on board. The airline collapsed in 2001 after being sold to private investors and Air France for $69 million, with debts of $500 million (many of which accumulated when the CFA franc col- lapsed in the 1990s). Governance issues are also commonly cited as a cause for the fall. When the airline finally ceased operating, there were a reported 4,200 employees, with only seven aircraft flying. Besides African destinations, the airline also flew to the Middle East, Europe, and the United States. Air Afrique's collapse removed a capacity of nearly 5 billion seat-km in 2001, similar in magnitude to the carrying capacity of Kenya Airways. Source: Bofinger 2009. of the collapse of Air Afrique and Air Nigeria, several nations surround- ing the BAG countries have experienced negative growth (see map 4.3). The lack of development in these countries, all of which have air traffic of less than 1 million passengers per year, is cause for concern. The North African market: Developing a hub. Two airlines, Royal Air Maroc and the slightly larger EgyptAir, carry 81 percent of the air traffic between Sub-Saharan Africa and North Africa. Three other North African airlines--Air Afrigiya, Air Algérie, and TunisAir--serve the remaining 19 percent. EgyptAir dominates along the east, and its route to and from Sudan accounts for nearly a fifth of all north-south travel. Royal Air Maroc predominantly serves the western side of the continent. The annual growth rate of traffic between North Africa and Sub-Saharan Africa Airports and Air Transport: Policies for Growth 157 Table 4.5 Top 14 Airports in Africa Serving International Travel within Africa Airport Estimated seats, Overall Country City/airport identification 2007 (thousands) percentage South Africa Johannesburg JNB 5,742 20.0 Kenya Nairobi NBO 2,901 10.1 Ethiopia Addis Ababa ADD 1,706 6.0 Nigeria Lagos LOS 1,157 4.0 Senegal Dakar DKR 986 3.4 Zambia Lusaka LUN 959 3.4 Uganda Entebbe EBB 954 3.3 Zimbabwe Harare HRE 828 2.9 Ghana Accra ACC 813 2.8 Namibia Windhoek WDH 791 2.8 Tanzania Dar es Salaam DAR 749 2.6 Côte d'Ivoire Abidjan ABJ 717 2.5 Mauritius Mauritius MRU 544 1.9 Angola Luanda LAD 484 1.7 Source: Bofinger 2009 (based on data from Seabury ADG). Map 4.3 Regional Growth Zones in Seats Offered, (all Travel) Good Growth (20% or more since 2001) Very High Growth (40% or more since 2001) Highest Growth (65% or more since 2001) Negative Growth (-20% or less since 2001) Source: Bofinger 2009 (based on data provided by Seabury ADG). 158 Africa's Transport Infrastructure exceeded 18 percent from 2001 to 2007 and reached almost 26 percent between 2004 and 2007. There are new routes between 45 country pairs, 17 more than in 2001. Of these, 41 have a single-carrier monopoly, including all of the new ones. The new routes are primarily with Morocco and Libya. Morocco is also an important hub for travel within Sub-Saharan Africa. The Libyan carrier Air Afrigiya, a relatively recent (2001) market entrant, offers a similar network to that of Royal Air Maroc. The develop- ing hub system in North Africa is therefore remedying the lack of a strong Sub-Saharan carrier on the west side of the continent. Domestic air transport: Growing despite national flag carrier protection. The number of seats offered for domestic air travel within Africa grew by more than 12 percent annually between 2001 and 2004. Growth varies widely among countries, however, depending on topology, population density, per capita gross national income, and the size of the tourist mar- ket (see appendix 4g for detailed country figures). South Africa (which accounts for 72.5 percent of all domestic services), Nigeria (accounting for another 10 percent), and Mozambique have expe- rienced the majority of growth in domestic air transport in the region. In fact, excluding these countries, domestic air service in the region declined by nearly 1 percent between 2001 and 2004. Overall, the number of inter- connected country pairs in Africa fell by 229 between 2001 and 2007, pri- marily a result of the collapse of major regional carriers (including Air Afrique and Ghana Airways). In most cases, the state carrier is the only provider of domestic service in Africa. Of the 286 routes with service in 2007, only 54 were served by more than one provider (appendix 4h). Occasionally, however, flag carriers subcontract less heavily traveled routes to private operators. For example, Air Malawi, which has scheduled flights on the Lilongwe­Blantyre route, will sometimes hire a small operator with a single-engine aircraft for flights with low load factors. South Africa and Tanzania have the most competition for domestic air travel in the region. Only the most heavily trafficked routes in South Africa have more than one service provider. In contrast, each of Tanzania's 17 domes- tic routes was served by more than one carrier as of 2007. Their com- petitiveness may now be affected by the continued problems of the flag carrier, Air Tanzania.9 The critical factor is not the actual number of airlines operating in any particular market but whether market entry is free enough to ensure efficient service. Airports and Air Transport: Policies for Growth 159 Market Structure: Stable, but Changing in Composition A small number of airlines dominate air travel in Sub-Saharan Africa. In 2007, 15 carriers provided 59.1 percent of the total seats in all markets (including the intercontinental market) in Sub-Saharan Africa, although that figure was down from 63.9 percent in 2001. In particular, SAA's mar- ket share fell from roughly 16 percent in 2001 to 14 percent in November 2007. British Airways also lost market share. On the other hand, Emirates, Ethiopian Airlines, and Qatar Airways are growing. Emirates has shown the greatest increase in capacity, from 960,000 seats in 2004 to over 3.6 million in 2004. It now has an almost 3 percent market share. Comair, an established South African airline with franchise agree- ments with British Airways, has also grown significantly. The market share of the top 15 carriers in an African market with a total seat capac- ity of 130 million seats and 319 billion seat-km as of 2007 is shown in table 4.6. The overall market is divided evenly between African and non-African carriers. (More information on market concentration can be found in appendix 4h.) International transport within Africa: Liberalization increasing concen- tration. As of 2007, 15 airlines accounted for over 82 percent of all pas- senger seats offered for international travel within Africa. The top three Table 4.6 Top 15 Airlines Overall in the African Passenger Market Estimated Market Market total seat-km, share, share, Rank Airline 2007 (millions) 2001 (%) 2007 (%) 1 SAA 34,112 15.7 13.8 2 Air France 22,707 7.7 7.6 3 EgyptAir 21,636 7.0 5.4 4 British Airways PLC 17,150 9.7 4.4 5 Emirates 14,504 1.1 4.1 6 Royal Air Maroc 13,772 3.4 4.0 7 Ethiopian Airlines Enterprise 12,493 2.1 3.9 8 Kenya Airways 11,602 2.4 2.9 9 KLM 10,688 3.4 2.8 10 Air Mauritius 8,598 3.3 2.5 11 Deutsche Lufthansa AG 7,676 2.5 1.8 12 Air Algérie 5,851 2.1 1.7 13 Virgin Atlantic Airways Limited 5,171 1.4 1.5 14 Tunisair 5,035 1.9 1.4 15 Qatar Airways (WLL) 4,623 0.2 1.3 Source: Bofinger 2009 (based on data provided by Seabury ADG). 160 Africa's Transport Infrastructure carriers--SAA, Ethiopian Airlines, and Kenya Airways--accounted for over 57 percent (see table 4.7). The number of carriers providing inter- national service fluctuated between 67 and 78 between 2001 and 2007. In 2007, 76 carriers served 206 country pairs, down from 238 country pairs in 2001. The decline in the number of country pairs with service accompanied an increase in market concentration by dominant players: 16 of the top 60 routes were served by only one carrier in 2007, up from 10 in 2001. Market concentration increased even more in the rest of the market; 50 routes were complete monopolies, up from 24 in 2001. On the other hand, 25 of the routes served by a monopoly carrier did not exist in 2001, reflecting the willingness of airlines to risk serving a new country pair. Ethiopian Airlines and Kenya Airways were dominant in these new markets. The 206 country pairs with service in 2007 accommodated an esti- mated capacity of 14.3 million passenger seats. SAA, Kenya Airways, and Ethiopian Airlines dominated 30 of the top 60 city pairs (as opposed to country pairs), which accounted for 80 percent of total capacity. Traffic among South Africa, Sudan, and Nigeria was among the Table 4.7 Top 15 Airlines Providing International Service within Africa Seat-km, Seat-km, Seat-km, Annual Annual 2001 2004 2007 growth, growth, Airline (millions) (millions) (millions) 2001­07 (%) 2004­07 (%) SAA 4,113 5,292 4,784 2.6 ­1.7 Ethiopian Airlines Enterprise 1,335 2,119 4,235 21.2 12.2 Kenya Airways 1,780 2,366 4,163 15.2 9.9 Air Mauritius 488 545 730 6.9 5.0 Delta Air Lines Inc. -- -- 639 -- -- Virgin Nigeria -- -- 598 -- -- Air Namibia 336 523 564 9.0 1.3 Zambian Airways 63 14 559 44.0 85.3 Air Senegal International 131 417 442 22.5 1.0 Airlink (ex South African Airlink) -- 201 406 -- 12.4 TAAG Angola Airlines 368 391 405 1.6 0.6 Bellview Airlines Ltd. 87 220 399 28.8 10.4 Air Zimbabwe (Pvt) Ltd. 402 175 383 ­0.8 13.9 Comair Ltd. -- 291 366 -- 3.9 Nationwide Airlines (Pty) Ltd. 31 117 263 43.1 14.4 Source: Bofinger 2009 (based on data provided by Seabury ADG). Note: -- = not available. Airports and Air Transport: Policies for Growth 161 fastest growing in the region. (More information on city pairs can be found in appendix 4c.) Monopoly: A hazard in low-density markets. The total number of pas- senger seats on routes served by one carrier (the monopoly market) grew by 6 percent annually from 2001 to 2007. Ethiopian Airlines is by far the largest monopoly carrier, serving 45 percent of the monopoly market, or nearly 1.2 million seats. Kenya Airways, at 22 percent, is a distant second. For comparison, SAA serves only about 1 percent of the monopoly mar- ket. It appears that Ethiopian Airlines has sought to expand into markets in which it can dominate. Its service in the monopoly market grew from a mere 327,400 seats in 2001 to its 2007 figure of 1.2 million, an annual growth rate of 27 percent. Furthermore, of the 21 country pairs for which Ethiopian Airlines is the sole carrier, only 6 did not exist in 2001, and 2 others were also served by a competitor that later abandoned the route. Kenya Airways seems to have followed a similar strategy at a smaller scale (though with higher growth rates), often driving out other operators by utilizing competitive advantages such as its well-developed hub system, newer fleet, greater ability to schedule capacity economically (with more types of aircraft available to meet demand), and the resources necessary to enter a price war if needed. Conventional methods of measuring market concentration indicate that service between country pairs tends to be oligopolistic, which is com- mon in low-density markets. National flag carriers: Still protected. Of the 53 states in the whole African continent, 25 have a national flag carrier in which the state has at least a 51 percent share. There are two main groups of flag car- riers: the three behemoths (Kenyan Airways, SAA, and Ethiopian Airlines); and the rest, most of which run large operating deficits. The high costs of fuel, maintenance, and insurance in Africa contribute to high operating costs for carriers. Many also serve very limited markets. Although the behemoths may run as separate corporate units, all three are primarily state owned. State ownership therefore cannot be solely responsible for the distressing financial condition of many flag carriers. The issue is rather that of small national market size and cor- respondingly low aircraft utilization levels. The outcome is also affected by the ambitiousness of the flag carrier in deciding how large a network to serve and also how well the fleet suits this size. The weaker flag carriers typically serve small domestic markets, which 162 Africa's Transport Infrastructure they try to subsidize from profits on international routes protected from competition by restricting the licensed air-service capacity of competitors. In fact, international routes could often be served more cheaply by larger international airlines, and smaller markets could be better served by small, private regional airlines. Attempting to priva- tize instead of liquidating flag carriers often leads to even larger sus- tained losses as countries pour good money after bad in support of essentially noncommercial operations. Fleet Composition: Younger and Better Adapted to Market Size Carriers in Africa operate a wide range of aircraft (300 different equip- ment codes are recorded). Table 4.8 lists the main types of aircraft in use, categorized by age group and origin ("Eastern" refers to the former Soviet Union) and by type (a combination of size and range). Two recent trends are apparent. First, the fleet became much younger between 2001 and 2007, with aircraft built in the 1980s and later accounting for 75 percent of travel in 2007--up from only 50 percent in 2001 (see figure 4.3 and appendix 4i for details on aircraft age). The second trend is in aircraft type, with a move away from wide-body and large jets toward smaller jets such as the Boeing 737 and Airbus 319 (see figure 4.4 and appendix 4j for country details). Between 2001 and 2007, the share of city jets in large international markets stabilized at over 60 percent, but increased substantially in midsize markets, from 34 percent to 52 percent. In small markets, the share of commuter propeller aircraft increased from 33 percent to 40 percent. According to the declining use of wide-body aircraft, it appears that route lengths shortened, since wide-bodied aircraft are economically used only on longer flights (table 4.9). Table 4.8 Breakdown of Aircraft Age for Analysis Age rating Aircraft type Western, very old vintage DC3 and similar; not in use in scheduled service Western, very old 1960s­70s, includes 727s, 737-100s, and similar Western, old 1970s­80s; 737 later series, early 747s Western, somewhat recent 1980s­90s (for example, Boeing 757) Western, recent The newest aircraft, generally from the mid-1990s onward Eastern Former USSR vintage; not large role overall Source: Bofinger 2009 (based on data provided by Seabury ADG). Airports and Air Transport: Policies for Growth 163 Figure 4.3 Trends in Aircraft Age, 2001­07 a. Seat-kilometers flown by aircraft age, 2001 unknown Western, very old vintage Western, very old Western, age unknown Western, very Eastern, age old to old unknown Western, old Western, recent Western, Western, somewhat old to somewhat recent to recent recent Western, somewhat recent b. Seat-kilometers flown by aircraft age, 2007 Western, very old unknown Western, very old vintage Western, very Western, age old to old unknown Western, old Western, old to somewhat Eastern, age recent unknown Western, somewhat recent Western, recent Western, somewhat recent to recent Source: Bofinger 2009. Service Quality: The Regional Hub as a Counter to Diminished Direct Connectivity It is still difficult to travel by air between African countries; often a con- nection in North Africa or Europe is required. Analysis of inter-African connectivity reveals a decline in international city-pair connections for 164 Africa's Transport Infrastructure Figure 4.4 Trends in Aircraft Type, 2001­07 a. Seat-kilometers by aircraft size, 2001 general aviation vintage propeller unknown eastern, undetermined wide-body jet commercial propeller large jet commercial jet city jet b. Seat-kilometers by aircraft size, 2007 eastern, undetermined general aviation unknown commercial propeller wide-body jet commercial jet large jet city jet Source: Bofinger 2009. Table 4.9 Breakdown of Aircraft Type by Market Size International Eastern built flights aircraft, Overall national one week in General Commuter Commuter Wide-body jet unknown Year market size November aviation (%) prop (%) jet (%) City jet (%) Large jet (%) (%) type (%) 2001 > 5 million 6,236 .. 13 1 65 0 20 1 > 1 million 2,169 .. 27 1 34 5 34 1 < 1 million 3,081 0.04 33 2 38 2 20 1 2007 > 5 million 10,638 .. 14 7 61 1 17 0 > 1 million 3,363 .. 17 5 52 2 22 1 < 1 million 3,167 .. 40 3 39 3 11 4 Source: Bofinger 2009 (based on data provided by Seabury ADG). Note: . . = negligible quantity. 165 166 Africa's Transport Infrastructure almost half of the countries between 2001 and 2007. This decline accel- erated between 2004 and 2007, affecting such diverse countries as the Central African Republic (only one international flight per week in November 2007), Chad, Eritrea, Mauritania, and the Seychelles. Worst hit by declining connectivity has been a group of countries sur- rounding Nigeria in West and Central Africa, together with the smaller markets of Côte d'Ivoire and Ghana, both with fewer than 1 million pas- sengers per year. This group includes the landlocked countries of Mali and Niger (along with the Central African Republic and Chad, which were already mentioned); it also includes coastal countries with smaller markets, such as Benin, Cameroon, the Republic of Congo, The Gambia, Gabon, and Togo. Air transport is financially unsustainable in these coun- tries, especially as national flag carriers rely heavily on Boeing 737­type jets that are too large and too expensive to operate in limited markets. The use of commuter propeller aircraft has increased slightly on inter- national routes in these markets. While this increase is a step forward, service quality could be much improved by further reexamination of fleet composition, flight frequency, and routing. For example, West Africa lacks a regional hub of intercontinental travel (North Africa, East Africa, and southern Africa all have one). One suggestion, therefore, has been to explore the development of a hub in Lagos. Turboprop-type transport aircraft--such as the Fokker 50, the ATR 42-300, and the Bombardier Dash-8 Q400--could expand the range of the hub by serving surround- ing countries (Bofinger 2009). This scenario is illustrated in map 4.4. The inner circle presents the range of an ATR 42-300, about 1,100 km. The middle range of roughly 2,000 km represents the range of a standard Fokker 50, while the outer ring, with a radius of 2,500 km, shows the range of a newer Bombardier Dash-8 Q400. Using the Fokker 50, the southern range of the hub would extend to Luanda, Angola. Even using the shorter-range ATR, the hub could service at least eight countries. Private operators such as Precision Air in Tanzania have been particularly successful in developing shorter routes with turboprop aircraft. A central hub would have several other advantages. By a better match of aircraft type to market demand, it would make load factors more sus- tainable per aircraft and thus enhance regional travel. Concentration of traffic through a regional hub might provide service to several countries, including those with little other traffic. Permitting operators fifth- and sixth-freedom rights (under the so-called freedoms of the air) to carry passengers between two countries outside their own is a vital prerequisite Airports and Air Transport: Policies for Growth 167 Map 4.4 African Countries Potentially Served by Commuter-Style Turboprop Aircraft Using a Hub in Lagos Source: Bofinger 2009. for the introduction of such a hub-and-spoke system.10 Implementation of the Yamoussoukro Decision (see below) provides for this. Airfares: Simply Structured but Expensive Flying to and within Africa is generally considered more expensive than flying to or within Europe or the United States, although this is difficult to show statistically because of the complex fare structures of the more developed markets, where fares change from minute to minute, depend- ing on load factors of individual flights. But there is little doubt that fares in general are higher per kilometer in Africa. Pricing schemes in Africa are not complex. Airfares were examined for a representative sample of routes: 23 international routes within Africa, 29 intercontinental routes, and 21 domestic routes. Because many 168 Africa's Transport Infrastructure domestic routes are not well advertised, domestic airfares were most dif- ficult to sample: only 13 price points were found. Standard booking Web sites were then used to determine the lowest-cost flights. Figure 4.5 shows the price of tickets per nautical mile for flights of various lengths, based on the samples. Air travel within Africa appears to be considerably more expensive per mile flown than intercontinental travel, especially on routes of less than 4,000 km. This result reflects the larger markets and higher competitiveness of the intercontinental routes. Economic Regulation and the Yamoussoukro Process Two common beliefs have limited entry into the air transport market of African countries. First is the belief that if services become too competi- tive after deregulation, then routes with low traffic volumes that are not economically viable will drop out of the system, isolating some parts of the country. Second is the belief that a national flag carrier, if owned and operated by the government and given sufficient market dominance, can support services on less viable--but socially necessary--routes with rev- enues from more profitable routes. As a result, each country protects its routes and allows airlines from other countries to enter its market only in exchange for a similar allowance. Much of the world has relaxed or removed such strict regulations. In the United States, deregulation has had several positive effects: increased price competition, the disappearance of weaker carriers, the more efficient Figure 4.5 Pricing of Flights within Africa versus Intercontinental Flights, Kilometers Flown 1.80 1.60 1.40 1.20 US$ per km 1.00 0.80 0.60 0.40 0.20 - - 2 4 6 8 10 12 14 distance (thousand km) international intercontinental domestic between SSA and NA Source: Bofinger 2009 (based on data collected by the World Bank). Airports and Air Transport: Policies for Growth 169 arrangement of routes, and the development of hub-and-spoke systems. In Europe, it has led to the rise of new low-cost carriers. The Yamoussoukro process marked the beginning of Africa's march toward liberalization. On October 17, 1988, the ministers responsible for civil aviation in the African states met in Yamoussoukro, Côte d'Ivoire, to propose a new African air transport policy. The result of that meeting was the Yamoussoukro Declaration. Although it sought the gradual elimina- tion of traffic restrictions, the declaration's primary goal was improved cooperation among African air carriers to allow them to better compete with non-African carriers. The declaration prompted the United Nations Economic Commission for Africa to initiate another conference of air trans- port ministers in Yamoussoukro, which resulted in the historic agreement on Pan-African liberalization of air services--the 1999 Yamoussoukro Decision (YD) (Schlumberger 2008). The YD had several main objectives: the gradual liberalization of entry into the African market, including both scheduled and nonscheduled air services; the abolishment of limits on the capacity and frequency of inter- national air services within Africa; the protection of universal traffic rights up to the fifth freedom; and freedom of operators to set fares. A monitor- ing body would supervise and implement the decision. Signatory states were obliged to ensure fair opportunity to compete in their air transport sectors, with plans for an African air transport executing agency to ensure fair competition. Special attention was given to improving air transport safety. Under the decision, all airlines were obliged to meet the standards defined by the ICAO, and states were asked to comply with established civil aviation safety and security standards and practices. Although most African states are bound to the YD,11 in practice it is implemented by regional economic organizations, rather than the Pan-African body of the African Union. The extent to which the YD has been implemented varies, as do the effects of implementation. The monitoring body has met only a few times, and competition rules and arbitration procedures are still pending. In 2007, at the Third African Union Conference of Ministers Responsible for Air Transport, the role of the executing agency was assigned to the African Civil Aviation Commission (an institution of the African Union). This agency has not yet proven effective in formulating and enforcing general rules and regulations governing competition. In contrast, opera- tional implementation has been much more productive. Countries in all subregions now have greater freedom to negotiate bilateral agreements (table 4.10). 170 Table 4.10 Air Service Liberalization among Regional Groupings in Africa General status of YD Status of air services Pecentage of fifth- and Community Members implementation liberalization seventh-freedom flights Banjul Accord Group (BAG) The Gambia, Ghana, Principles of the YD agreed Up to the fifth freedom 43 Guinea, Mali, Nigeria, upon in a multilateral air granted, tariffs free, Sierra Leone, Togo service agreement and capacity and frequency open Economic Community of Cameroon, Central African Principles of the YD agreed Up to the fifth freedom 28 Central African States Republic, Gabon, Equatorial upon in an air transport granted, tariffs free, and (CEMAC) Guinea, Republic of Congo, program; still some minor capacity and frequency Chad restrictions open; maximum two carriers per state permitted to participate Common Market for Most states of East and Full liberalization agreed Pending; once applied, 14 Eastern and Southern southern Africa, except upon, but implementation operators permitted to Africa (COMESA) Tanzania, South Africa, pending until a joint serve any destination Botswana, and Lesotho competition authority (all freedoms); tariffs established and capacity/frequency to be free. East African Community Tanzania, Uganda, Kenya Directive issued by EAC Air services not liberalized; 16 (EAC) Council with goal of amendment of bilateral amending bilateral accords still pending accords among EAC states to conform to the YD Southern African Develop- Most countries south of No steps taken toward No liberalization within 6 ment Community (SADC) Tanzania implementation, even SADC initiated though the civil aviation policy includes gradual liberalization of air services within SADC West African Economic and Benin, Burkina Faso, Côte YD fully implemented within All freedoms, including 44 Monetary Union (WAEMU) d'Ivoire, Guinea-Bissau, the WAEMU cabotage, granted; tariffs Mali, Niger, Senegal, Togo liberalized Source: Bofinger 2009. 171 172 Africa's Transport Infrastructure In West Africa, the West African Economic and Monetary Union (WAEMU) fully implemented the YD, and even went beyond it to guar- antee cabotage rights. The BAG, also in West Africa, agreed to a multilat- eral air service agreement that was fully compatible with the decision.12 In Central Africa, the CEMAC implemented all the necessary legislative and regulatory elements to comply with the provisions of the YD. In East Africa and in southern Africa, the COMESA achieved the most progress, but implementation is still pending, conditional on the establishment of a joint competition authority. The East African Community has chosen the effective strategy of directing countries to amend their bilateral agree- ments to conform to the decision, but the agreement has not yet been signed. The SADC, in the south, has progressed least; the dominant posi- tion of South Africa appears to be the main obstacle to the implementa- tion of the decision. Overall, two-thirds of the air transport service in Africa is now liberalized. An examination of the nationality of carriers flying international routes within a region clearly shows the effects of liberalization. Despite a net loss in the number of city pairs and country pairs served directly, there has been a significant increase in the percentage of routes served by carriers not based in either the country of origin or the country of destination (see table 4.11). The two main exceptions to this trend, the CEMAC and WAEMU regions, are explained by the failure of Air Afrique, which was not considered a national carrier in any country. Its failure therefore reduced the number of routes carried by a nonnational carrier. In many cases, extraregional African carriers (such as an East African carrier traveling between two countries within WAEMU) are replacing the capacity of the lost carriers (Air Afrique, Air Gabon, Ghana Airways, and Nigeria Airways), while European carriers once flying similar routes (such as Air France) have almost completely disappeared in the region. This suggests that the larger and healthier carriers are consolidating serv- ices in these markets. While there is some anecdotal evidence that fares for flights from a carrier's home country to another country (third- and fourth-freedom operations) have declined because of the YD, there is no solid analysis using long-term fare data to support that proposition. As with most efforts to liberalize air transport, countries wishing to protect unhealthy flag carriers have resisted implementing the YD (Schlumberger 2010). As in other regions, a few very large flag carriers dominate air transport in Africa (South African Airways, Ethiopian Airlines, and Kenya Airways). Smaller flag carriers, which sometimes con- sist of one or two aircraft in Africa, fly any profitable routes between their Table 4.11 Percentage of Flights between Country Pairs Served by Airlines Not Based in Either Country Seats Country pairs City pairs As of Net change As of Net change Annual growth Annual growth November from November from Regional community Total 2007 2001­7 (%) 2004­7 (%) 2007 February 2001 2007 February 2001 Arab Maghreb Union (AMU) 1,294,189 4.55 8.65 9 -- 14 2 Banjul Accord Group (BAG) 568,306 0.32 13.87 13 -- 15 1 Economic Community of Central African States (CEMAC) 152,984 ­18.88 ­35.58 6 ­6 9 ­9 Common Market for Eastern and Southern Africa (COMESA) 4,484,675 7.12 17.66 49 ­4 71 ­3 East African Community (EAC) 1,751,811 2.02 5.81 9 1 18 ­2 Southern African Development Community (SADC) 5,663,632 4.27 10.00 34 ­4 72 5 West African Economic and Monetary Union (WAEMU) 763,472 ­5.42 ­5.56 20 ­2 21 ­3 Source: Bofinger 2009 (based on data provided by Seabury ADG). Note: -- = not available. Flights are international flights within each region. Except for the Arab Maghreb Union, which is not part of the YD, all countries have shown an increased market proportion of these airlines between 2004 and 2007. The data for 2001 are skewed because several regional airlines with large market shares, such as Air Afrique, collapsed. 173 174 Africa's Transport Infrastructure country and outside hubs, while attempting to sustain an otherwise unprofitable network. Liberalization allows the dominant carrier based in the regional hub to compete on profitable routes, and thus the small flag carrier becomes completely unsustainable. Efforts to protect a flag carrier deprive passengers of choices, which usually results in higher prices and lower service quality. Safety: Achilles' Heel African airlines, although they carried only 4.5 percent of total air traffic, were responsible for a quarter of all fatal air transport accidents world- wide in 2007. The African Airlines Association argues that this is a result of aging fleets: nearly a third of the region's 750 aircraft are over 20 years old. Soviet-built aircraft are still common in certain countries. But their danger relative to Western aircraft seems more a question of vintage rather than origin. While most accidents in 2006 involved old Soviet-built turboprop aircraft, more recent crashes have mainly involved Western- built aircraft. Data from around the world suggest that properly main- tained and operated Soviet aircraft are as safe as Western aircraft from the same vintage. Meanwhile, inquiries by the U.S. National Transportation Safety Board have highlighted several cases in which poor pilot training and assessment contributed to aircraft accidents in Africa. The IATA identifies poor regulatory oversight as the top threat to air safety in Africa, followed by inadequate safety management systems. Only Cape Verde, Ethiopia, and South Africa meet international stan- dards for safety. Similarly, results from the ICAO's Universal Safety Oversight Program reveal that safety performance throughout Africa is very poor. For example, as figure 4.6 shows, West and Central Africa and East and southern Africa perform worse than the world average in every critical measure of safety implementation (in most cases by a factor of 2). A cross-sectional analysis of the region reveals that these deficiencies are highly correlated with accident rates, suggesting that institutional failings explain a large part of Africa's accident record. (More indicators of poor air safety can be found in appendix 4k.) The high accident rate in the air sector in Africa has caught the atten- tion of donor countries, development institutions, and industry-related associations and organizations. The ICAO, the U.S. Department of Transportation, the Industry Safety Strategy Group (ISSG, formed by Boeing, Airbus, and several associations), AviAssist of the Netherlands, the French Civil Aviation Authority, and the World Bank have all imple- mented programs to improve air safety in the region. For example, the Airports and Air Transport: Policies for Growth 175 Figure 4.6 ICAO Analysis of Safety Implementation 70 60 percentage not implemented 50 40 30 20 10 0 primary specific CAA structure technical qualified licensing and continued resolution of aviation operating and safety guidance technical certification surveillance safety issues legislation regulations oversight material personnel obligations obligations functions critical element global East and Southern Africa West and Central Africa Source: ICAO 2007. Department of Transportation's Safe Skies for Africa Program has helped East Africa establish a new regional safety oversight organization. The ICAO is helping to set up three projects under its Cooperative Development of Operational Safety and Continuing Airworthiness Program for the WAEMU, CEMAC, and BAG countries, which may even- tually lead to additional regional flight-safety oversight agencies. Africa's regional associations are also pooling resources to address safety issues. For example, the African and Malgache Civil Aviation Authorities, an association of 15 civil aviation directors general from West and Central Africa, was established in 2001 to further cooperation in the supervision of aviation safety in the region. The ISSG's program has established an overarching set of goals for the sector, which serves to coordinate donor and other aid activity. The ICAO, with assistance from the World Bank, is creating a central repository and database for projects related to air transport, which will be mapped to other metrics, such as the ISSG's program. Progress in the region cannot yet be discerned if one looks only at acci- dent statistics. Yet there have clearly been successes, such as the creation of a more independent CAA in Nigeria. The ongoing global recession and the potential for rising fuel costs are poised to limit growth in Africa's air transport sector. Continued improvement in the region's aviation safety is therefore more critical than ever to the success of the industry. 176 Africa's Transport Infrastructure The Way Forward Five policy objectives appear to be of the highest importance for meeting the challenges of growth in the African air transport sector. Priority 1. Improve safety oversight Three policy commitments are critical to improving the region's poor air safety record: · Political commitment to the autonomy of the national safety oversight organizations, which may require both new laws and independent auditing · Budgetary commitment of national governments to support regional safety oversight organizations and enable them to pay sufficiently competitive salaries to retain technically competent personnel, to be shared throughout the region · Commitment to regional pooling of information and resources devoted to the supervision of aviation safety. Priority 2. Invest in maintenance rather than construction of airport facilities In general, Africa's runways are sufficient to meet demand. At the same time, some terminals are congested, and both airside and landside facili- ties are often outdated. Priorities for the region's investment strategy are therefore as follows: · Discourage investment for replacing airports with new ones. · Focus investment in maintaining and upgrading infrastructure, including runways, taxiways and aprons, terminals, and landside access to airports. · Invest in smarter and less expensive ATC and navigational infrastructure. · Encourage private sector participation in landside investments; in partic- ular, landside service provisions (such as check-in, baggage-handling, and cargo terminal operations) could be outsourced to specialized firms. Priority 3. Avoid spending to support unprofitable flag carriers With only a few exceptions, flag carriers are highly unprofitable. But gov- ernments often continue to subsidize flag carriers for fear of losing unprofitable domestic routes. Countries should therefore pursue the fol- lowing strategies: · Liquidate perennially unprofitable flag carriers. · Liberalize important routes, opening them to both foreign and domes- tic carriers. Airports and Air Transport: Policies for Growth 177 · If necessary, supplement sustainable services with competitively ten- dered net cost contracts for economically unviable routes. Priority 4. Reform the financing of air traffic infrastructure In most countries, the two main agencies concerned with air traffic could be made more effective by adopting a more commercial attitude to the revenues yielded by the services they provide: · CAAs, which rely on fees for funding, should be assured of at least a predictable share of revenues from overflying. · Airport authorities, which are inherently profitable because of their monopoly position, should receive all revenues that they generate, but should also adequately maintain runways and terminals; surpluses could be taxed if necessary. Priority 5. Further liberalize the air market Implementation of the YD has improved international connectivity in countries that have lost carriers since 2004. The increase in fifth- and sixth-freedom operations conducted by Ethiopian Airlines, Kenya Airways, and SAA is particularly indicative of progress. Progress has been slower in countries that still protect their flag carriers. The following actions can hasten progress: · Concentrate on developing regional agreements, which have been very successful when implemented. · Immediately apply the liberalized policies to all domestic markets. Notes 1. The main source document for this chapter is Bofinger (2009). The discussion of service liberalization is based on Schlumberger (2007, 2008). 2. http://www.aircraft-charter-world.com. A list of airports was composed by combining this Web site's list of airports for every country in Africa. Data sources for this chapter are described in appendix 4a. 3. The BAG comprises Cape Verde, The Gambia, Ghana, Liberia, Nigeria, and Sierra Leone. 4. Radar vectoring is the provision of navigational guidance to aircraft in the form of specific headings based on the use of radar. Radar separation is applied by a controller observing that the radar returns from the two aircraft are a cer- tain minimum horizontal distance from each other, as observed on a suitably calibrated radar system. Secondary radar is a system that not only detects and 178 Africa's Transport Infrastructure measures the position of aircraft, but also automatically asks for additional information such as identity and altitude. 5. ILS is a ground-based instrument system that provides precision guidance to an aircraft approaching and landing on a runway, using a combination of radio signals and, in many cases, high-intensity lighting arrays to enable safe landing in poor-visibility conditions. 6. This can occur where service networks with relatively short connections, as in modern hub-and-spoke arrangements, also have relatively low frequencies. If one airport cannot accept aircraft because of bad weather, the entire connec- tion schedule is disturbed. 7. Jeppesen (also known as Jeppesen Sanderson) is a subsidiary of Boeing Commercial Airplanes that specializes in aeronautical charting and navigation services, flight planning, pilot supplies, and aviation training. 8. The Safe Skies for Africa Program was inaugurated by President Clinton on April 1, 1998. The goals of the program were to improve safety, security, and air navigation in Africa. The first eight countries selected to participate were Angola, Cameroon, Cape Verde, Côte d'Ivoire, Kenya, Mali, Tanzania, and Zimbabwe. Djibouti, Namibia, and Uganda were added in June 2003. Over the intervening years, Safe Skies assistance has been expanded to help regions share personnel and other resources to make up for a lack of qualified tech- nical personnel in individual states. 9. Originally established as a state enterprise after the breakup of East African Airways in 1977, Air Tanzania has had a checkered history. In 2002, South African Airways signed an agreement with the government of Tanzania to be the strategic partner of the Air Tanzania Corporation, and purchased a 49 percent stake in the company. The new airline was launched in March 2003 with Dar es Salaam as a hub. After four loss-making years, the govern- ment bought out the SAA interest, terminated the agreement, and relaunched Air Tanzania in October 2007. In December 2008, the Tanzanian Civil Aviation Authority withdrew its air operator certificate and the com- pany was banned by the IATA. The promised reestablishment of service seems unlikely to be sustainable. 10. The eight "freedoms of the air" are the focus of international regulation of air transport. The first and second freedoms allow aircraft to overfly a foreign country or to land for refueling. The third and fourth freedoms are commer- cial freedoms to carry passengers from a carrier's home country to another or vice versa. The fifth to seventh freedoms concern the rights to carry passen- gers between two foreign countries--as an extension of a flight from the home country (fifth), via a stop in the home country (sixth), or without ongo- ing service to the home base (seventh). The eighth freedom is the right to carry traffic between two points in a foreign country. Airports and Air Transport: Policies for Growth 179 11. The Yamoussoukro Decision was agreed to on the basis of the Abuja Treaty, which set up the African Economic Community. Thus, only the 44 states that have signed and formally ratified the Abuja Treaty are parties to the YD. Nonsignatory states were Djibouti, Equatorial Guinea, Eritrea, Gabon, Madagascar, Mauritania, Somalia, South Africa, and Swaziland. 12. The regional associations overlap considerably. Both the WAEMU and the BAG countries are members of the larger Economic Community of West African States (ECOWAS). There is also considerable overlap of membership in the COMESA, East African Community (EAC), and SADC. References ACI (Airports Council International). 2008. Airport Economics Survey, 2008. Geneva: ACI. ADPI. 2008. "Analyse Economique et Financière des Capacités de Développement des Aéroports du Mali: Projet d'amélioration de l'Aéroport International de Bamako Selnou." ADPI Designers and Planners, Mons, France. Bofinger, H. 2009. "Air Transport: Challenges to Growth." Africa Infrastructure Country Diagnostic Background Paper 16, World Bank, Washington, DC. Button, K. 2008. "Air Transportation Infrastructure in Developing Countries: Privatization and Deregulation." In Aviation Infrastructure Performance: A Study in Comparative Political Economy, ed. Clifford Winston and Gines de Rus, 193­221. Washington, DC: Brookings Institution Press. ICAO (International Civil Aviation Organization). 2007. "Universal Safety Oversight Audit Program Report." ICAO, Montreal. Osborne, A. 2007. "BAA Face Penalties If London Airports Cut Investment." Telegraph Media, October 5. Schlumberger, C. 2007. "Emerging Issues for Air Navigation Services: A Challenge for Developing Countries." Paper presented at "Aviation Safety, Security and the Environment" conference, McGill University, Montreal, September 15­17. ------. 2008. "The Implementation of the Yamoussoukro Decision." McGill Institute of Aerospace Law, McGill University, Montreal. ------. 2010. Open Skies for Africa: Implementing the Yamoussoukro Decision. Washington, DC: World Bank. World Bank. 2009. Public-Private Infrastructure Advisory Facility Database. World Bank, Washington, DC. http://ppi.worldbank.org/explore/ppi_explore Region.aspx?RegionID=1. CHAPTER 5 Ports and Shipping: Moving toward Modern Management Structures Africa has many ports, most of which are small by world standards.1 Few can accommodate the largest ships. In general, African ports are poorly equipped, have low productivity, and are unprepared for the rapidly unfolding changes in global trade and shipping patterns. While they are moving slowly from public ownership and operation to the landlord port- management model, they still lag behind in the development of modern port-management structures compared to ports in other regions. Coping with Rapidly Changing Trade Patterns Transport demands are a function of trade patterns, which are changing rapidly. Africa's market share in the 30 most significant (in terms of traded value) non-oil exports fell from an average of 20.8 percent in the 1960s to less than 10.0 percent in the 1990s. In many countries, this decline was due to poor economic performance triggered by the upheavals of the post- colonial period, described in chapter 1. Over the past decade, Africa's mar- ket has recovered slightly in absolute terms, led by the growth of oil exports from West Africa, but not in terms of world market share. Changing trade patterns will alter the types of traffic transported. 181 182 Africa's Transport Infrastructure Spatial traffic patterns are also changing. Trade has declined with Europe and has grown with East Asia. Trade with North America has risen as a consequence of the U.S. search for oil security. China, India, the Republic of Korea, and Malaysia are also quickly developing an interest in Africa's energy products. Container Traffic: Fast Growth from a Low Base Container traffic in Africa, with the exception of South Africa, is still at an early stage of system development. In 2005, African ports handled a com- bined 8.6 million twenty-foot equivalent units (TEUs),2 of which Durban handled nearly 2.0 million. The three main South African ports together handled over 35 percent of this traffic--3 million TEUs. Although con- tainer traffic in Africa is growing rapidly, it started from a very low base (figure 5.1). Container traffic in West Africa, for example, has grown at an average annual rate of 14.7 percent--the highest of all African regions-- but still accounts for less than 1 percent of container traffic worldwide. Traffic in East Africa, which is the second-fastest-growing subregion, is heavily concentrated in Mombasa (5 percent of the total for Africa). And in West Africa, five ports each handle more than 350,000 TEUs. In south- ern Africa, traffic rose from 1.35 million TEUs in 1995 to 3.09 million in Figure 5.1 Container Trade Development Summary in the Study Ports, by Region, 1995 and 2005 5 15 15 average annual growth rate (%) 4,082 container volume (million TEU) 4 11 3,092 10 3 9 2 1,395 1,356 5 1,035 1 505 0 0 East Africa/ Southern Africa West Africa Indian Ocean container traffic volume by region 1995 2005 average annual growth rate Source: Mundy and Penfold 2009. Ports and Shipping: Moving toward Modern Management Structures 183 2005. As of 2005, all Africa accounted for little over 2 percent of container traffic worldwide. Durban is by far the largest port in the region. Its container volume in 2005 was 1.9 million TEUs, more than twice the volume of any other port. Conakry and Cape Town are the next largest, handling 753,827 and 690,895 TEUs in 2005, respectively. Abidjan comes in fourth with a container volume of 500,119 TEUs (figure 5.2). Several ports--among them the larger ports of Abidjan and Luanda and the smaller ports of Cotonou, Nacala, and Walvis Bay--experienced dramatic growth of around 30 percent every year between 1995 and 2005, compared with the 10­15 percent typical among other fast-growing ports worldwide. In the future, these African ports are likely to see accelerated growth in container traffic. Several important factors are influencing the growth of African con- tainer trade. First, increased stability and economic growth in the region has led to rising demand for manufactured goods mainly imported in containers. This is particularly true in countries that have benefited from surging petroleum revenues (Nigeria, for example). Second, the globalization of production has led to growing trade between Africa and Asia, including imports of Asian consumer goods. Third, global innovations have resulted in increased penetration of the container system into general cargo trade and have generated a cascade Figure 5.2 Top 10 Sub-Saharan African Ports (TEUs per year) 2.00 1.75 1.50 TEU (million) 1.25 1.00 0.75 0.50 0.25 0 r e Bay m r da th a an n y an ka rk s w aa ba be dj rb an na Da To ds al i om Du iza Ab Lu Co sS pe ar El M ch Ca rt Ri Da Po port Source: Mundy and Penfold 2009. 184 Africa's Transport Infrastructure effect, whereby major shipowners are gradually deploying larger and more modern container vessels in African trade. Finally, institutional reforms have allowed foreign expertise and investment to enter the port sector and have enabled public-private partnerships (PPPs) of var- ious forms, such as concessions and management contracts for container- handling facilities. But Africa is still at a relatively early stage in developing its container system. In 2006, the region had a population of 770.3 million and accounted for a total of 8.60 million TEUs in all its ports. For compari- son, North America had a population of 528.0 million and accounted for a total traffic volume of 50.10 million TEUs. Clearly, considerable poten- tial remains for container traffic volumes to increase in Africa. Constraints on Growth: Transit Arrangements Most landlocked countries have several outlets to the sea. For example, the five landlocked countries in West Africa have 15 transit corridors, while in southern Africa, Zambia alone has 5. In southern Africa, traffic often travels the longer route to the port of Durban as a result of the more liberal land transport and border arrangements on that route and the fre- quent sailings from the port. Table 5.1 identifies a number of regional ports and the markets from which they capture, or could capture, transit traffic. Container traffic development in Africa is presently constrained by transit arrangements in these corridors. More competition between corri- dors, as well as comprehensive corridor development programs (for example, the Maputo corridor), could reduce administrative blocks and allow goods to flow more freely. A revival of Africa's inland waterways could also play a constructive role (box 5.1). Most road and rail systems serving the corridor ports are in poor condition and do not support container transport (Harding, Palsson, and Raballand 2007). With a few exceptions (such as in South Africa), containers are stuffed and stripped close to the port of entry or departure. As a result, the volume of containerized traffic moving to landlocked countries across land borders is very low. In a recent study, Mundy and Penfold (2009) did not find much evidence that national or regional transport authorities have prioritized port devel- opment, or that port management is generally involved in developing links to land transport. Current practices impede development of both the ports and the economies they serve. Coordinated investments in ports, railways, and roads, along with other trade facilitation initiatives, could allow a number Table 5.1 Regional Ports and Their Transit Traffic Markets--Actual and Potential Country Port Transit traffic markets Volume, 2005 Comments East Africa Sudan Port Sudan, Suakin n.a. -- Potential to reach Central African Republic, Sudan, Rwanda, Uganda, Burundi, Chad by rail Djibouti Djibouti Ethiopia n.r. Well established Kenya Mombasa Uganda n.r. Well established Tanzania Dar es Salaam Burundi, Rwanda, Uganda, 50,000 TEU Established but constrained by capacity Zambia Mozambique Maputo, Beira Malawi, Zambia, Zimbabwe, n.r. Significant potential but volumes low South Africa, Swaziland Southern Namibia Walvis Bay Botswana, Zimbabwe, South Africa n.r. Connections offered but volumes low Africa South Africa Durban, Cape Town, Port Elizabeth Botswana, Namibia, Mozambique n.r. Significant capability by rail but volumes low West Africa Angola Luanda Congo, Dem. Rep.; Zambia; Zimbabwe n.a. No capability at present, but major rail refurbishment under way Congo, Dem. Rep. Matadi Central African Republic; Congo, Rep. n.a. No capability foreseen Congo, Rep. Pointe Noire Central African Republic 400 TEU No capability at present Cameroon Douala Burkina Faso, Central African n.r. Connections offered but volumes low Republic, Chad, Mali, Niger Cameroon Kribi Kribi-Kissangandi Development n.r. Under development Development Corridor is planned connecting Equatorial Guinea; north Gabon; Congo, Rep.; Congo, Dem. Rep. Benin Cotonou Niger n.k. Established but constrained by capacity Togo Lomé Burkina Faso, Mali, Niger n.r. Connections offered but volumes low Ghana Tema, Takoradi Mali, Niger, Burkina Faso n.r. Ghana Gateway Program exists but faces capacity issues Côte d'Ivoire Abidjan Burkina Faso, Mali, Niger n.r. Resumes after the war; agreements with rebels to secure road access Senegal Dakar Mauritania, Mali 33,400 TEU Established and with significant potential for further development 185 Source: Author's compilation of third-party industry sources and Mundy and Penfold 2009. Note: -- = not available; n.r. = not reported; n.a. = not applicable; n.k. = not known. 186 Africa's Transport Infrastructure Box 5.1 Inland Waterways: A Neglected Asset Inland waterway transport has historically been important for carrying primary product exports from landlocked countries but is now in decline. The three major lakes in East and Central Africa--Victoria, Tanganyika, and Malawi--once played an important role in transit and intraregional trade in the region. On Lake Victoria, in particular, waterway transport was linked to railheads at the inland ports of Kisumu (Kenya), Bell (Uganda), and Mwanza (Tanzania). The Ugandan and Kenyan lake operations were concessioned together with the railways in those countries, while in Tanzania, the lake services have been separated from the railways since the introduction of the Uganda and Kenya concessions. Only one service now operates on Lake Victoria, and some of the railway track leading to the ports is in poor state of repair, especially in Kenya. A similar story applies in West and Central Africa, where the Congo basin has a navigable network of 12,000 kilometers and covers nearly 4 million square kilo- meters in nine countries. In principle, the Congo system could be a very valuable resource in a multimodal transport network serving the region. In practice, how- ever, it suffers from outdated and insufficient infrastructure (as well as inadequate channel markings and maintenance, feeble regulation, and numerous nonphysi- cal barriers to movement) and plays an ever more marginal role as a mode of transport. Recognizing this untapped potential, in October 2005, the executive secretary of the Economic and Monetary Community of Central Africa (CEMAC) encouraged the governments of Cameroon, the Republic of Congo, the Demo- cratic Republic of Congo, and the Central African Republic to establish the Commission Internationale du Bassin Congo-Oubangui-Sangha to improve the physical and regulatory arrangements for inland navigation in the basin (CICOS 2007). A consultancy study has been undertaken to examine the cur- rent arrangements in the four participating countries, and to identify the steps that need to be taken to begin effective redevelopment. Source: Author. of the ports identified in table 5.2 to become more effective gateways for international trade, particularly containerized traffic. For example, the Maputo Corridor Development project comprises road, rail, border posts, port, and terminal facilities between Mozambique's port of Maputo and South Africa, Swaziland, and Zimbabwe, and it has significantly enhanced the attractiveness of the Maputo port. Table 5.2 Major African Ports and Their Transshipment Potential Traffic, in Country Port Transshipment Status as of 2010 TEUs (2005) East Africa Djibouti Djibouti Significant Existing capacity constraints to be relieved by opening of new terminal in 2009­10 60,000 Kenya Mombasa Substantial Role declining because of pressing capacity constraints 54,576 Tanzania Dar es Salaam Substantial Initially took business from Mombasa, but now experiencing capacity constraints also 40,000 Southern Africa South Africa Durban Substantial Natural center for southern Africa, but now facing capacity constraints 212,000 West Africa Nigeria Apapa Potential Improvements in port efficiency helping to realize potential, but capacity constraints remain an issue n.r. Cameroon Douala Potential No capacity constraints at present, but may emerge on the basis of new facilities under development n.k. Benin Cotonou Potential No capacity constraints at present, but may emerge the basis of on nearby new port development (Seme-Kpodji) n.k. Ghana Tema Significant Currently constrained by capacity, but new development plans exist 50,000 Côte d'Ivoire Abidjan Substantial Business suffering as a result of recent conflict 300,000 Senegal Dakar Substantial Currently constrained by capacity and low efficiency; recent concession may improve situation 80,000 Source: Mundy and Penfold 2009. Note: n.r. = not reported; n.k. = not known. 187 188 Africa's Transport Infrastructure Transport investment is not necessarily the main component of a trade development program. For example, the Ghana Gateway Program has three main components--development of a free trade zone, trade facilita- tion, and investment promotion--and involves seven implementing agen- cies in an effort to approach development in a comprehensive way. The first phase of this project involved a comprehensive set of reforms focus- ing on frontline institutions such as the Ghana Free Zones Board, Customs Excise and Preventive Services, Ghana Immigration Service, Ghana Ports and Harbors Board, and Ghana Civil Aviation Authority. Such relatively inexpensive reforms may produce greater benefits than expensive invest- ments in port capacity. Container Trade: Exceptionally Imbalanced Flows Most global container trade is imbalanced, with flows of loaded contain- ers in one direction exceeding those in the opposite direction. This imbal- ance has only become more pronounced with the proliferation of large-scale manufacturing facilities in China. In East Africa, West Africa, and southern Africa, imports account for the majority of loaded container movement (see figure 5.3). Moreover, most containers are exported empty. In 2005, the ratio of empty exported TEUs to loaded exported TEUs was 90:10 in West Africa; 80:20 in East Africa; 65:35 in Southern Africa; and 80:20 in Africa as a whole. For comparison, on other arterial container trade routes, such as transpacific or Asia­Europe, the ratio ranged from 30:70 to 40:60. Thus, the imbalance is worse in Africa than in most regions of the world. Container Transshipment Transshipment is common for most African ports, as the national and regional markets do not generate sufficient demand to justify a place on the itinerary of the major intercontinental shipping lines. These markets are therefore dependent on calls from smaller container vessels carrying cargo that has been transshipped at a larger port. To play a role in trans- shipment, ports should have deepwater and good container-handling per- formance, and be unencumbered by excessive bureaucracy. Ports that can also generate gateway cargo are even more desirable. Based on these cri- teria, all three maritime zones in Africa must improve significantly to effectively handle container transshipment. Regional transshipment ports on the East African coastline include Mombasa and Dar es Salaam. Both face capacity constraints likely to Ports and Shipping: Moving toward Modern Management Structures 189 Figure 5.3 The Imbalance of Sub-Saharan African Container Trade, 2005 3.5 3.0 2.5 TEU (million) 2.0 1.5 1.0 0.5 0 East Africa West Africa Southern Africa exported empty exported loaded imported loaded Source: Mundy and Penfold 2009. curtail their transshipment activity, at least in the short term, although both are moving to install new capacity. In contrast, at the Port of Doraleh in Djibouti, a new terminal designed with transshipment in mind and developed by DP World, is now operational.3 This terminal provides sig- nificant transshipment capacity for East Africa and the Indian Ocean and will compete with Aden and Jeddah (both also partly operated by DP World) for cargo from the main traffic lane through the Red Sea. The port may develop a large industrial area and could attract transshipment for Ethiopia and Sudan. In southern Africa, Durban is well established as the major con- tainer transshipment center, but it too has been struggling to keep pace with demand. There are plans for new facilities (such as the new Pier 1), but demand may still outstrip the capacity of these additions. A number of carriers that use Durban for transshipment are seeking alternate ports in the Indian Ocean islands, notably Mauritius, despite 190 Africa's Transport Infrastructure the extra sailing time required to service the region from Mauritius rather than Durban. There are also doubts about whether the cost of expanding capacity in Mauritius to accommodate transshipment traf- fic can be met with transshipment revenues, which are usually lower than those generated by gateway traffic. Elsewhere in South Africa, Transnet has opened a new terminal in Coega, designed to play a major role in transshipment operations, and has announced plans to develop more facilities in Richards Bay, which could offer alternatives to Durban. On the West African coast, Abidjan has enjoyed some success as a con- tainer transshipment port. In recent years, however, it has suffered because of internal strife in Côte d'Ivoire and disputes over the award of the concession of operating rights for the container terminal.4 There are plans for a number of new major hubs, but none is likely to be developed in the immediate future. Transshipment traffic for West Africa, unlike that for the east and south, is fed primarily from outside the region, via Algeciras and the new Tangier terminals. There is clearly potential to further develop container transshipment capacity in Africa. But prospects for expanded capacity may depend on initiative from a major line or consortia to drive hub development. General Cargo Traffic: Growing by Default General cargo traffic in Africa is growing more quickly than in other regions of the world, largely because the region is behind the rest of the world in the development of containerization. Figure 5.4 shows growth from 1995 to 2005, broken down by subregion. Although exports of agricultural products and raw materials are con- siderable at a number of ports in the region, the balance of trade for gen- eral cargo traffic is heavily weighted toward imports (though not as heavily as the container sector). Transit traffic to landlocked countries is predominantly general cargo. A positive economic outlook, even withstanding the global financial crisis, has increased revenues from oil production, and persistent con- straints on the container traffic system bode well for continued growth in general cargo traffic in the near future. Capacity is sufficient to meet demand, though governments must ensure that this continues to be the case. Cargo-handling performance, measured in terms of tonnes handled per man-hour or per gang-hour, is below par by world standards, however, and calls for improvement. Ports and Shipping: Moving toward Modern Management Structures 191 Figure 5.4 Development of General Cargo Traffic, 1995­2005 80 20 average annual growth rate (percent) 18 61.2 60 15 metric ton (million) 40 11 38.4 10 10 23.1 20 5 14.5 13.8 2.7 0 0 Southern Africa East Africa and West Africa Indian Ocean general cargo traffic volume by region 1995 2005 average annual growth rate Source: Mundy and Penfold 2009. Dry Bulk Traffic: Efficient Dedicated Terminals The broad spectrum of dry bulk traffic in Africa (table 5.3) comprises two categories: major bulk, consisting of commodities such as coal, iron ore, and grain, which tend to be moved in substantial volumes; and minor bulk, consisting of cement, aggregates, clay, and other commodities that tend to be moved in smaller volumes. Dry bulk traffic is sometimes han- dled at common-user general cargo facilities. Major flows--both in terms of volume and value (such as grain at Mombasa and coal from Richards Bay)--use industrial-style terminals that are often privately owned and do not always publicly report traffic volumes. For example, Richards Bay has one of the largest coal-exporting facilities in the world; but while the general cargo port is publicly owned, the coal- exporting facility is entirely private. Since major global interests con- trol the facility, port and shipping arrangements will likely conform to international standards. Liquid Bulk Traffic: A Well-Oiled Commercial Chain Liquid bulk traffic, predominantly oil, is a growing sector in Africa. The United States and the Asian countries have recently made significant Table 5.3 Dry Bulk Operations in Africa 192 Country Port Existing and planned dry bulk operations East Africa Kenya Mombasa The grain terminal in this port is under private sector operation. It was concessioned via an international tender and has operated very efficiently since the new consortia took over. Mozambique Maputo The port has recently set up a specialized ferrochrome terminal as a niche dry bulk operation, servicing exports of this commodity from South Africa. The port also operates coal export terminals. Madagascar Talarno A new port development is designed to facilitate exports from a new $350 million ilmenite mine being developed by Rio Tinto, by far the largest mine development undertaken in Madagascar for some time. Southern Africa South Africa Richards Bay The Richards Bay Coal Terminal is expanding its export capacity from 78 million to 92 million metric tonnes, at a cost of 1 billion rand, making it the biggest facility of its kind in the world. West Africa Cameroon Douala The master plan studies undertaken for the new port address how to exploit the country's natural resources, including dry bulk commodities such as bauxite, iron ore, cobalt, nickel, and rutile. Benin Cotonou The port is mainly used for the import of cereal and gypsum, but is currently operating beyond capacity. The new master plan incorporates an expansion at nearby Seme-Kpodji to accommodate growing volumes of these commodities. Togo Kpeme The privately operated port of Kpeme exports phosphate rock. Liberia Harper Carvalla Rubber Corporation plans to rehabilitate the port to provide required bulk export capacity. Buchanan Arcelor Mittal has agreed to increase its investment in Liberian ore production from $1.0 billion to $1.5 billion. The ore will be exported from Buchanan, where significant upgrade works are planned to port facilities. Ghana Takoradi A new master plan has been commissioned for Ghana's two ports of Tema and Takoradi. For Takoradi, this will focus on the development and concessioning of new dry bulk terminals. Sierra Leone Pepel The government of Sierra Leone has granted African Minerals the exclusive right, under a memorandum of understanding, to rehabilitate the port of Pepel. The intention is to refine and expand the port to achieve a throughput of 40 million tonnes per year (mt/yr). Guinea Kamsar Kamsar is now the scene of some significant proposed development as part of the so-called Guinea Alumina Project. New port facilities will include a dedicated alumina export terminal. Guinea-Bissau Angola has secured the right to develop a 3 mt/yr bauxite mine in Guinea-Bissau, and this will involve new port construction on the Buba River. Source: Mundy and Penfold 2009. Ports and Shipping: Moving toward Modern Management Structures 193 investments in the region to establish the necessary export platforms, including pipeline and shipping jetty facilities. Figure 5.5 gives an overview of oil production in Africa. The region has 11 net oil exporters: Nigeria, Angola, Equatorial Guinea, Sudan, the Republic of Congo, Gabon, Chad, Cameroon, Mauritania, Côte d'Ivoire, and the Democratic Republic of Congo. Nigeria and Angola are the dominant oil states, in terms of both production rates and proven reserves. Recently, Ghana and Benin have begun to produce oil, although not enough to satisfy their domestic needs. Africa is a significant supplier of oil to countries worldwide. It con- tributes more than 10.0 million barrels per day to the daily world output of 84.5 million barrels, accounting for almost 12 percent of the world oil supply and approximately 19 percent of U.S. net oil imports. The region's natural gas markets are also growing. Asian countries (including China, India, Korea, and Malaysia) are rapidly developing an interest in Africa's energy products. As a consequence, the energy sector accounted for a large part of the nearly fourfold increase in total trade between China and Africa in 2001­06. The United States has invested substantially in Africa's energy sector, reflecting its desire to decrease reliance on the Middle East by seeking Figure 5.5 Oil Production and Proven Reserves in Sub-Saharan Africa 3.0 40 35 2.5 proven reserves (Gb) 30 production (Mb/d) 2.0 25 1.5 20 15 1.0 10 0.5 5 0 0 ia la a n p. n ad on ia re . ep ne da bo r go an Re oi ge Ch o .R ui Su Iv er Ga rit An o, Ni m lG d' m au ng De Ca te ria M Co Cô o, to ng ua Co Eq country production 2005 proven reserves, January 2007 Source: U.S. Department of Energy 2006; Oil and Gas Journal 2006. Note: Gb = 109 billion barrels; Mb/d = million barrels per day. 194 Africa's Transport Infrastructure energy sources outside that region. The majority of U.S. energy invest- ment in Africa--over $11 billion and growing--is in the oil and gas sec- tors of Equatorial Guinea, Nigeria, and Angola. Oil companies from the United States are also heavily involved in oil production in Chad and in the $4 billion investment in the Chad-Cameroon pipeline project. U.S. companies are producing and looking for oil and gas in South Africa, and are also exploring opportunities for energy investment in Benin, Guinea- Bissau, Madagascar, and São Tomé and Príncipe. Recently, there has been a large oil find at sea close to the Liberian­Sierra Leone border. There are few financial impediments to putting the right export platforms in place. In the past, state-owned organizations and private interests have usually tailored the development of new export capacity (ports and shipping) for the energy sector to meet the needs of producers outside the mainstream sphere of port operations. Future projects will probably adhere to this model, which has worked well in the past. (See appendix 5a for more information on the development of traffic types.) Constraints on Economies of Scale Countries in Africa have long since abandoned their efforts to enforce the United Nations Liner Code.5 As a result, the African shipping mar- ket has been largely deregulated, allowing it to better integrate with the global market. Shipping systems have evolved in a global context, the largest con- tainer vessels now possessing capacities in excess of 13,000 TEUs. The average size of container vessels serving African ports is relatively small, under 3,000 TEUs, but it is steadily creeping up in line with port system improvements. The acquisition of regional operators by global players and the replacement of direct service to or from ports of origin by transship- ment through hubs have accelerated Africa's integration into the global liner network. For example, Maersk Line uses Salalah, Oman, as its hub for East African trade; for its West African trade it uses Tangier, Morocco, and Málaga and Algeciras, Spain. The constraints on economies of scale imposed by size limitations on vessel access to African ports raise the costs of shipping to Africa. So do port inefficiencies and inadequate links between ports and hinterlands. As a result, feeder services (particularly in East Africa) and regional liner services (in West Africa) will continue to be important. Costly delays are a problem in many ports. They are mostly a result of long processing times and poor handling in congested port areas rather than insufficient quay capacity. In 2006, a one-day delay cost a shipping Ports and Shipping: Moving toward Modern Management Structures 195 line $35,000 for a 2,200-TEU vessel, with proportionally higher costs for larger ships. Shipping lines have responded by introducing congestion charges that range from the equivalent of $30 for a 20-foot container in Dakar to $360 in Tema in 2006. Where customs authorities allow the transport of boxes under bond (allowing customs clearance to be moved to an off-dock location), some ports have developed off-dock terminals to relieve container yard congestion. The global financial crisis has temporarily relieved the pressure on prices and capacity, although not without adverse effects. A global decrease in maritime transport has left Africa with an oversupply of shipping capacity. Nearly 600 container vessels were without work in September 2009, and there was a significant increase in the number scrapped. The major shipping lines suffered large losses in the first six months of 2009, and more were expected to follow in the short term. In sum, shipping to or from Africa remains expensive. This is not due to any inherent inefficiency or lack of competition in shipping operations (factors that explain high cost in other trade lanes). Rather it is due to the absence of the large and concentrated flows necessary to capitalize on economies of scale in deep-sea shipping. Also at fault are the high costs of land distribution, particularly to landlocked countries. The Institutional and Regulatory Framework The institutional and regulatory framework for the port sector has three components: port facility planning, customs arrangements, and regulation of port management (Bell and Bichou 2007). The treatment of each of these functions differs substantially among countries. Planning and Reform: A Challenge for Government New master plans for the port sector were either recently introduced or are under development in Namibia, South Africa, Tanzania, and 10 countries in West Africa. Although planning focuses on the devel- opment of facilities and physical capacity, a number of strategic issues must also be addressed. Africa can support only a limited number of major regional hubs. Competition is already intense among ports in East Africa, and regional collaboration seems unlikely. A port must do more than invest in capac- ity to become a hub; it must also be able to offer low handling costs and have fairly high cargo potential in its local market. In addition, it should facilitate transit traffic by developing the main trade corridors from the 196 Africa's Transport Infrastructure port to the landlocked hinterland. The failure of governments and port authorities in Africa to speed international, intermodal transport by streamlining customs and other formalities at border crossings has stifled trade, slowed inland movement, and driven up prices. Although private operators are responsible for developing their logistics chains, they rely on governments to coordinate port, customs, and inland transport arrangements. Governments have several responsibilities in strategic port planning: they must establish the respective roles of the public and private sectors, outline methods for attracting and selecting private partners, and identify the technologies and management arrangements needed to develop state- of-the-art ports. These responsibilities require the involvement of the international private sector, particularly in the container terminal busi- ness. Countries with congested city ports or limited water depth at quay- side or in access channels will need to decide whether to rehabilitate existing ports or develop new ones. Developments in the deep-sea shipping markets may also force countries to relocate their ports. For example, to be economical, a ship- ping route between Asia and Latin America requires vessels with capac- ities of 6,000 TEUs or more. Any such service would benefit greatly from a port of call in South Africa. But Durban has insufficient capac- ity, and the container terminal in development in Cape Town is too far from the industrial core of the country in Gauteng to be a strong hub port. Developing Richards Bay, which has deep water and ample space, might be a better option, and Transnet has reported plans to undertake development there. Customs: Seeking Simplification and Automation Customs and other procedures associated with the movement of goods across international borders increase the time and cost of transport and impede its flexibility (De Wulf and Sokol 2004). For example, in almost half the African countries surveyed in a recent study, it took more than one week to clear goods through the major port, compared with one day or less in many developing countries in other parts of the world. Several factors were identified as contributing to this poor performance (McTiernan 2006). First, there is little mutual trust and understanding between customs authorities and the business community. The private sector sees customs administrators as primarily concerned with collecting revenue. In all countries but Togo, frontline customs officers are thought to be unaware Ports and Shipping: Moving toward Modern Management Structures 197 of, or unsympathetic to, the business headaches caused by their perform- ance. This is particularly pronounced in countries where officers dress in military-style uniform. In Kenya, Tanzania, Uganda, and Zambia, customs officers are part of the national revenue authority and do not dress like the military. Meanwhile, officers in Côte d'Ivoire, Ghana, and Nigeria dis- play their respective rank on military uniforms and are perceived to be more autocratic. Second, many agencies are involved, making coordination difficult. Some bodies--such as those certifying food, drugs, and agricultural products--operate alongside but separate from customs. While port authorities are separate, their processes and procedures are intrinsically linked with those of customs. Ministries of trade and commerce some- times slow customs processes by starting initiatives that, alongside existing procedures, just add to the bureaucratic load. Third, dissemination of information is poor. Some customs services (for example, in Côte d'Ivoire and Senegal) make good use of the Internet. Others have elaborate Web sites that have not been updated for months, even years. Especially where the use of clearing agents is required or commonplace, it can be difficult for businesses to identify the cause of long residence times, whether related to customs procedures or those of other agencies. Fourth, there is little standardization of procedures and documentation. Other than the Southern African Customs Union and the Trans-Kalahari corridor of the Southern African Development Community, both of which are adopting a single clearance document, trade blocs seem to have little influence on customs practices. For example, the members of the East African Community, all of which also belong to other blocs, have adopted different practices in key areas. Kenya uses a version of the Senegalese Trade-X system as the software for its clearance process, while Tanzania and Uganda run ASYCUDA++.6 Tanzania operates a full inspection program subcontracted to a third party. Kenya relies on importers to secure certificates of conformance with the Kenya Bureau of Standards. Uganda has no inspection or certification program. Fifth, corruption is a serious problem. In all the countries surveyed by the Business Action for Improving Customs Administrations in Africa (see McTiernan 2006), corruption was believed to be present at an indi- vidual level rather than systemwide or through organized syndicates. In a commonplace scenario, individual customs officers are bribed to pass undervalued goods. Customs and other officials require bribes to reward them for using their discretion to speed the clearance process. The scope 198 Africa's Transport Infrastructure for this type of bribery is reduced by automation. The more automation, the less opportunity there is for individual manipulation. The Democratic Republic of Congo, even with few systems and institutions, thus suffers from substantial low-level corruption. Nigeria, which has a large customs service and detailed processes, but no electronic system, is reported to experience much individual corruption on customs transactions. While the example of Rwanda demonstrates that it is possible to eliminate cor- ruption even in a very poor country, the incentive for individual corrup- tion is likely to remain substantial while incomes remain low. For an approach to the problem, a number of changes have been already suggested (McTiernan 2006) and a handbook has been produced on reform options (De Wulf and Sokol 2005). With development of a consultative arrangement among stakeholders (including businesses), a basis might be found for rewarding a record of compliance with "fast track" systems; at virtually no risk to revenue, the burden of many proce- dures and costs could be lifted from those businesses that represent a low risk for compliance evasion. The different agency requirements could be integrated by the establishment of a "one-stop shop" for documentation and clearance. Acceptance of a common electronic documentation sys- tem such as ASYCUDA would be central to that goal, though customs officers should still be able to manually overrule a "green channel" desig- nation automatically generated by the ASYCUDA++ system. But it is no coincidence that those countries that do not have electronic clearance in place are invariably seen as the most inefficient and corrupt. Simplification and automation are probably the most powerful practical weapons against corruption, alongside relevant training. Some progress has already been made. Customs reform and modern- ization programs are currently under way in many countries in Africa, although in general, reform appears less advanced in West Africa than in East or southern Africa. In southern Africa, the Southern African Development Community is proceeding with the implementation of shared documentation and progressively reduced internal tariffs. It has not advanced as far as the Common Market for Eastern and Southern Africa, where internal tariffs have already been eliminated. A common authorized economic operator model is to be implemented as a top priority by five countries of the East African Community (Burundi, Kenya, Rwanda, Tanzania, and Uganda), with mutual recognition among them. When implemented, the model will be one of the first of its kind in the world and will likely benefit the business community. Ports and Shipping: Moving toward Modern Management Structures 199 Port Management Models There are three main port management models (World Bank 2007). Service port. In this model, the port authority offers the complete range of services required for the functioning of the port. The port owns, maintains, and operates every available asset (fixed and mobile), and cargo-handling activities are executed by labor employed directly by the port authority. Most service ports are government owned, though there also are some private service ports usually dedicated to a single major shipper or commodity (such as the Richards Bay Coal Terminal in South Africa). Landlord port. This model is characterized by a mix of private and public sector functions. The public sector port authority acts as a regu- latory body and a landlord, outsourcing port operations (especially cargo handling) and leasing infrastructure to private operating compa- nies and industries. The private operators employ the dock labor, and provide and maintain their own on-quay equipment and buildings. Either the public or the private sector can be responsible for pilotage, towage, and line handling. Whole port concession. Under this model, the public sector hands over complete responsibility for port management and operations to the private sector for a fixed number of years. Djibouti is the main example in Africa. Africa lags far behind most of the world in introducing institutional reforms to the port sector. Ghana and Nigeria are the only countries in Africa that have comprehensively embraced the landlord port model, although others have arrangements that incorporate some of its ele- ments. Container terminal activities are very attractive to the private sector and are therefore usually concessioned first. Lease contracts and build-operate-transfer concessions (where the concessionaire builds the facilities and operates them during the period of the concession but transfers them to the authority at the end of the concession) are the most common forms of concession (World Bank 2007). Such agreements have tapped the expertise of international terminal opera- tors, and more African countries are now adopting some aspects of the landlord model (table 5.4). The region has seen very little progress toward the establishment of independent port regulators, although experience shows good concession contracts can be self-regulating. (The institutional characteristics of the major ports are tabulated in appendix 5b.) 200 Africa's Transport Infrastructure Table 5.4 Port Management Models by Region Port management model Partial Whole Region Service port landlord Landlord port concession East Africa and Indian Sudan Tanzania n.a. Djibouti Ocean Kenya Madagascar Mozambique Southern Africa Namibia n.a. n.a. n.a. South Africa West Africa Congo, Dem. Rep. Angola Nigeria n.a. Congo, Rep. Gabon Ghana Equatorial Guinea Togo Benin Cameroon Guinea-Bissau Côte d'Ivoire Cape Verde Liberia Sierra Leone Guinea Gambia Senegal Source: Mundy and Penfold 2009. Note: n.a. = not applicable. Private Sector Participation: Increasing Steadily Changes in port management models have been associated with increased private sector participation in the funding of African port infrastructure (figure 5.6). Twenty-six ports spread over 19 African countries reported 42 major private sector transactions in recent years (table 5.5). Most are conces- sion contracts, and a large number of these are associated with the com- prehensive port reforms in Nigeria. The management contract for the Mombasa container terminal and original concessions for the Gabon ports of Owendo and Gentil have been canceled. The transactions listed in table 5.5 include private sector investment commitments of $1.3 billion. Of this total, about 62 percent relates to development of container terminals, 32 percent to multipurpose termi- nals, and very little to bulk facilities. Ports in Nigeria attracted 55 percent of total private sector invest- ment commitments in the sector; by far the largest commitment--over $300 million spread across six terminal concession contracts--involved the port of Apapa in Lagos (box 5.2). These transactions have generated substantial royalty payments from concessionaires to African govern- ments--$1.7 billion in total, including over $1.0 billion from the Apapa concession. Ports and Shipping: Moving toward Modern Management Structures 201 Figure 5.6 Evolution of Private Sector Participation in African Ports 30 number of transactions 25 20 15 10 5 0 93 94 95 96 97 98 99 00 01 02 03 04 05 06 19 19 19 19 19 19 19 20 20 20 20 20 20 20 management or lease contract concession contract greenfield projects Source: World Bank 2008. Private sector participation varies greatly among cargo sectors. Concessions have been most prevalent in the container terminal field (table 5.6). In some cases, the port authority or government agency has been reluctant to completely divest its assets, reportedly due to an inade- quate rate of return on investment, and instead has chosen to remain as a part owner of the port or terminal operating company. This occurred in the Mozambique whole port concessions and the Tema container terminal concession in Ghana. Several of the global terminal operators--including APM Terminals, DP World, and International Container Terminal Services Inc. (ICTSI)--participate in many of the region's port concessions. Some are facing financial problems as a result of the global economic downturn and have announced that they will decrease the pursuit of new ventures (although companies such as ICTSI continue to expand). This general trend may not be the case in West Africa, however, where global terminal operators seem to be competing fiercely for new concessions. Private sec- tor operation is standard in liquid bulk cargo facilities, but only one-third of the more than 50 dry bulk ports reviewed in Mundy and Penfold (2009) have concessioned any services to private operators, and none appears as a major concession initiative in table 5.6. Some container terminal concessions have been controversial. For example, rival bidders brought legal challenges after the Dakar and Luanda container terminal concessions were awarded to APM Terminals 202 Table 5.5 Private Sector Transactions for All Port Sectors in Africa Royalty Number of payments to Investment Number of canceled government in facilities Countries Ports transactions transactions (US$ millions) (US$ millions) Management or Cameroon, Kenya, Douala, Mombasa, lease contract Mozambique Maputo 4 1 0 0 Concession Angola, Comoros, Luanda, Mutsamudu, contract Equatorial Guinea, Luba, Owendo, Gabon, Ghana, Tema, Toamasina, Madagascar, Beira, Maputo, Mozambique, Quelimane, Apapa, Nigeria, Sudan, Calibar, Harcourt, Tanzania Lilypond, Onne, Warri, Tin Can, Juba 32 0 1,366 1,052 Greenfield Côte d'Ivoire, Abidjan, Luba, Tema, projects Equatorial Mombasa, Freeport Guinea, Ghana, Kenya, Mauritius 6 0 316 236 Total 42 1 1,682 1,288 Source: World Bank 2008. Ports and Shipping: Moving toward Modern Management Structures 203 Box 5.2 The Nigeria Port Concessions The Nigerian port system once comprised traditional service ports with ineffi- cient central management. The government of Nigeria initiated major reforms beginning in late 2004 to introduce the landlord model and concession over 20 terminals to the private sector. This was one of the most ambitious and far- reaching port reforms undertaken in Africa or elsewhere. The reforms have already benefited the ports. Operational benefits include improved turnaround time for ships and cargo, improved cargo-handling performance, and improved cargo and personnel security measures. Prior to reform, the Nigerian Ports Authority was subsidized by the government of Nigeria. Today, the reformed and downsized authority is largely self-funding. Private operators are scheduled to invest in excess of $500 million in port development, and will pay more than $5 billion to the government in rental or royalty fees. Shipping lines also reduced their congestion surcharges from $800 to $100 per container within a few months of concessioning, which saved the Nigerian economy $310 million and reduced excessive charges and corruption. Source: Mundy and Penfold 2009. and DP World, respectively. The concessioning process needs to be mod- ified to minimize the effects of corruption and influence. One model for reform is presented by Nigeria's Bureau of Public Enterprises, which hired experienced independent consultants to design the concessioning of the country's marine terminal facilities and oversee its implementa- tion. The involvement of the bureau ensured that concessioning was undertaken at arm's length from established port authority and govern- ment interests. Regulatory Arrangements In most countries, port regulation is undertaken either by a central min- istry for transport or by a port authority (table 5.7). These institutions fulfill a variety of economic and technical regulatory functions. All are involved in either the day-to-day management and operation of ports or, at the very least, formulating policy regarding port activities. Only South Africa has an independent port regulator, which monitors and enforces the compliance of the National Ports Authority 204 Table 5.6 Major Institutional Reform Initiatives Implemented or Planned for African Ports, as of 2009 Region Country Port Year Type Duration Contractor East Africa Djibouti Djibouti 2000 Whole port concession, plus 20 years DP World development of greenfield high-capacity container terminal Tanzania Dar es Salaam 2000 Container terminal concession n.k. ICTSI, transferred to Hutchison Mozambique Beira 1999 Whole port concession 25 years Cornelder (67%), CFM (33%) Mozambique Maputo and 2003 Whole port concession 15 years with MD and HC resold to DP World Matola option of consortium including 10-year extension Mozambique MIPS, Maputo 1996 Container terminal 15 years, DP World extended to 2113 Mozambique Nacala 2005 Whole port concession 15 years CDN Mozambique Quelimane 2005 Whole port concession 25 years -- Madagascar Toamasina 2005 Container terminal concession 20 years ICTSI West Africa Angola Luanda 2007 Container terminal concession 20 years APM Terminals Gabon Owendo 2007 Multipurpose terminal 25 years Gabon Port Management concession (Portek) Gabon Gentil 2007 Multipurpose terminal 25 years Gabon Port Management concession (Portek) Togo Lomé Planned New container terminal 25 years MSC (51%), Lomé (49%) concession Cameroon Douala 2004 Container terminal concession 15 years Consortium including APM Terminals Nigeria Multiple 2007 Container terminal and other Various Diverse investors at Apapa/ concessions Lagos container terminal APM Terminals Ghana Tema 2003 Container terminal concession 25 years Consortium including APM Terminals Côte d'Ivoire Abidjan 2005 Container terminal concession n.r. Consortium including APM Terminals Liberia Monrovia Planned Concession of multipurpose -- Subject to tender process facilities Sierra Leone Freetown Planned Concession of multipurpose -- Subject to tender process facilities Guinea Conakry Recent Concession of container terminal 25 years GETMA International Guinea Kamsar Planned Private sector concession in the n.k. n.r. Guinea Alumina Project Guinea-Bissau Bissau Planned Concession of cargo-handling n.k Not yet determined facilities Senegal Dakar 2006 Container terminal concession DP World Cape Verde Mindelo Planned Concession of cargo-handling n.k Not yet determined facilities 205 Source: Mundy and Penfold 2009. Note: n.r. = not reported; n.k. = not known; n.a. = not applicable; -- = not available. CDN = Corredor de Desenvolvimento do Norte; CFM = Caminhos de Ferro de Moçambique; ICTSI = International Container Terminal Services Inc.; MD and HC = Mersey Docks and Harbour Company; MIPS = Maputo International Port Services; MSC = Mediterranean Shipping Company. 206 Africa's Transport Infrastructure Table 5.7 National Institutions Responsible for Port Regulation in Africa, by Type Port planning Port operating Independent Region Transport ministry authority authority agency East Africa and Djibouti, Kenya Mozambique, Sudan, Tanzania n.a. Indian Ocean Madagascar Southern Africa n.a. n.a. Namibia South Africa West Africa Angola; Benin; Congo, Cameroon n.a. Nigeria Dem. Rep.; Congo Rep.; (planned) Gabon; Equatorial Guinea; Togo; Ghana; Côte d'Ivoire; Liberia; Sierra Leone; Guinea; Guinea-Bissau; Senegal; Cape Verde Source: Mundy and Penfold 2009. Note: n.a.= not applicable. with the National Ports Authority Act and serves as the appeals body for complaints and grievances lodged against the National Ports Authority. Nigeria also plans to establish an independent regulator. Infrastructure Development The capacity of a port may be increased by increases in quay length or standing area (usually referred to as infrastructure) or by improvements in the amount or quality of loading and unloading equipment (referred to as superstructure). Capacity and Demand: Some Ports Approaching Full Utilization Estimates of port capacity and demand for a sample of ports (selected for their accuracy in reporting these figures) suggest that several ports have either reached or are close to reaching full capacity (table 5.8). In East Africa, double-digit growth in the container sector has pushed the ports of Mombasa and Dar es Salaam to their capacity lim- its, and Sudan, Mombasa, and Dar es Salaam are reaching their limits in the dry bulk sector. In southern Africa, Durban has faced challenges in its efforts to expand container capacity to meet demand. And in West Africa, the ports of Luanda and Tema have insufficient container capac- ity, while Luanda, Douala, and Tema are struggling to handle overall cargo throughput. By late 2006, African port capacity was estimated to be at 80 percent utilization overall and forecasted to remain at that level through 2010 Table 5.8 Annual Port Capacity and Current Demand, Selected Ports Region Port Current capacity (as of 2006) Demand (as of 2006 unless otherwise stated) East Africa Port Sudan 9 million (m) metric tonnes 7.5 m metric tonnes 400,000 TEU Mombasa 17 m metric tonnes 15.87 m metric tonnes (2005) 400,000+ TEU 358,762 TEU (2005) Dar es Salaam 4.1 m metric tonnes dry cargo 3.8 m metric tonnes dry cargo (2005) 6.0 m metric tonnes liquid cargo 2.0 m metric tonnes liquid cargo (2005) 250,000 TEU 350,000 TEU Southern Africa Cape Town 1.0­1.2 m metric tonnes general cargo 559,602 metric tonnes general cargo (2005) 1.0 m TEU/yr containers 690,895 TEU (2005) Dry bulk and liquid capacity not reported Durban 1.3+ m TEU (being expanded) 1.9 m TEU (2005) General cargo, dry bulk, and liquid capacity 7.1 m metric tonnes general cargo not reported 5.5 m metric tonnes dry bulk 7.1 m metric tonnes liquid cargo West Africa Luanda 400,000 TEU 377,208 TEU 4 m metric tonnes per year 3.2 m metric tonnes general cargo 2.2 m metric tonnes roll-on roll-off (ro-ro) Douala 270,000 TEU 190,700 TEU (2005) 6.5 m metric tonnes general cargo 5.8 m metric tonnes general cargo 1.0 m metric tonnes liquid cargo Liquid cargo demand not reported Cotonou 200,000 TEU 158,201 TEU (2005) 2.5 m metric tonnes general cargo 1.0 metric tonnes (2005) 0.7 m metric tonnes liquid cargo Liquid cargo demand not reported Tema 350,000­400,000 TEU 420,000 TEU 8­9 m metric tonnes 7.9 m metric tonnes, all noncontainerized cargo Abidjan Not reported, but said to be approaching full 500,000 TEU+ (2007) 207 capacity limits by 2010 21.4 m metric tonnes noncontainerized cargo (2007) Source: Mundy and Penfold 2009. 208 Africa's Transport Infrastructure (Drewry Shipping Consultants 2006, 2009). Port congestion in some parts of Africa (particularly in Nigeria and in Mombasa and Dar es Salaam) was close to reaching critical levels, although the pressure has been reduced because of the global financial crisis. Cargo-handling rates are below international standards at all of the ports mentioned above. For example, Durban Container Terminal man- ages only about 17 container moves per hour, compared with the inter- national norm of 25 to 30. To remedy the situation, Durban is upgrading cranes at existing facilities and has introduced new systems at Pier 1. Although superstructure and infrastructure are separated in South Africa, the fact that both are controlled by the publicly owned Transnet has clearly compromised their proper development. Improvements are required if South Africa's ports are to realize their potential as interna- tionally important transshipment centers for southern Africa. Obstacles lie in the organization, provision, and management of equipment and handling space, as much as in basic quay capacity. Solutions include insti- tutional reform and mobilization of private sector capabilities in port service management as well as public sector investment. Port Development: Works in Progress A number of countries in the region--including those still designing their master plans--have either proposed or already begun major port devel- opments (table 5.9). Smaller schemes span a diverse range of activities, such as dredging new channels and maintaining existing ones, improving navigation sys- tems, rehabilitating berths, setting up new cargo-handling equipment, installing information technology (IT) systems, and purchasing security systems. The steadily increasing presence of the private sector in frontline cargo-handling operations provides added impetus for development. Cargo-Handling Systems: A Very Mixed Bag Cargo-handling systems in the ports vary widely, ranging from outdated to state of the art. (Available facilities for all ports are reported in appen- dix 5c.) A number of major ports--including Banjul, Dakar, Monrovia, Onne, Pointe Noire, and Port Harcourt--lack quay crane equipment and rely on ships' gear for cargo handling, while others have outmoded con- tainer gantry cranes.7 The continued use of outdated equipment limits port productivity and contributes to the region's capacity shortage. New investment in cargo-handling equipment, rather than in quay capacity, is thus needed to allow African ports to fulfill their potential. Ports and Shipping: Moving toward Modern Management Structures 209 Table 5.9 Principal New Port Developments in Africa Region Country Project East Africa Sudan Proposed: introduction of new container and other cargo capacity at Suakin port Djibouti Actual: new container terminal scheduled to commence operations in 2009, offering major new transshipment capacity Kenya Proposed: second container terminal for Mombasa Southern Africa South Africa Actual: development of new Pier 1 container terminal facility (Durban); opening of new port of Ngqura for container-handling operations; ongoing expansion of coal export capacity at Richards Bay; and major expansion of container terminal capacity at Cape Town through expansion of the terminal footprint, equipment, and other system developments Mozambique Proposed: upgrade of container capacity in Maputo, deepening and further improvement of port of Beira Madagascar Proposed: new port development at Talanaro to facilitate ilmenite exports West Africa Congo, Rep. In progress: Extension of Point Noire breakwater and container terminal plus the deepening of Terminal G Gabon Proposed: Expansion of Owendo quay and improvement of connecting road system Cameroon Actual: dredging of Douala port Proposed: new container terminal development Benin Proposed: new port development at Seme-Kpodji Togo Proposed: new container terminal development Ghana Proposed: extensive new port development at Takoradi Côte d'Ivoire Proposed: major capacity expansion at the Port of Abidjan Liberia Proposed: multipurpose terminal development Guinea Proposed: extension of the container terminal Senegal Proposed: addition of major new container port capacity Source: Mundy and Penfold 2009. Several ports have already benefited from public sector investment in new quayside container gantry cranes and landside-handling systems (notably by the Kenya Ports Authority in Mombasa and the National Ports Authority in South Africa). But the private sector, which has a grow- ing presence in container handling, continues to account for the most investment in quayside and landside container-handling and IT systems. The gradual transfer of the financial burden of such investments from governments to the private sector has had positive results. This trend is also evident in the dry bulk sector, where the industrial character of many terminals is particularly conducive to private sector investment. 210 Africa's Transport Infrastructure New equipment must be introduced in the proper context to realize its full benefit. For example, modern container gantries demand quay structures that can accept high loadings, and container terminal stacking areas must have sufficient structural strength to support multiple stacked containers. Modern container-handling systems demand not only the req- uisite equipment but also a high degree of organization by port manage- ment, proper IT support, and an increased emphasis on reducing damage to both personnel and property. Finally, proper environmental protection will assume much greater importance with the installation of new sys- tems for dry bulk handling, particularly in facilities that handle commodi- ties where spillage and contamination can occur. The importance of effective equipment maintenance and support systems has become clear in some state-owned ports that continue to deliver unsatisfactory per- formance despite the addition of new equipment. Deficient soft infrastructure also represents a significant opera- tional bottleneck at many ports in the region. In particular, customs procedures are often outdated or subject to corruption, and fre- quently delay cargo clearance (and have even prompted temporary port closures). Other obstacles include lack of IT-supported manage- ment, information, and communication systems both within the port area and down the supply chain. While upgrading physical equipment has been the focus of many port development projects, improving soft infrastructure is equally important to the continued development of the sector in Africa. Safety and Security Arrangements Safety at sea, and particularly in congested port areas, requires com- mitment to discipline in navigation. Since, by its nature, the shipping business is global, the rules must also be global. The function of the International Maritime Organization (IMO) has historically been to set these rules. Where the IMO's rules relate to the conditions governing crew employment, the International Labour Organization (ILO) has also been involved. In principle, responsibility for implementation of the IMO and ILO standards rests with the countries in which the vessels are registered. But since many vessels carry flags of convenience,8 appropriate action by the country of registration cannot be relied on. For that reason many of the IMO's most important technical conventions contain provisions for ships to be inspected at their location of business. "Port state con- trol" is the inspection of foreign ships in national ports to verify that the Ports and Shipping: Moving toward Modern Management Structures 211 condition of the ship and its equipment comply with the requirements of international regulations and that the ship is manned and operated in compliance with these rules. While the primary responsibility for ships' standards still formally rests with the flag state, port state control provides a safety net to catch substandard ships. A ship going to a port in one country will normally visit other countries in the region before embarking on its return voy- age, and it is to everybody's advantage if inspections can be closely coor- dinated. The IMO has therefore encouraged the establishment of regional organizations and agreements to govern port state control. Regional memoranda of understanding have been signed, including one for West and Central Africa. The African Union has also begun to take a more active role in this area. African experts on maritime security and safety met in April 2010 under the auspices of the African Union to consider an African Integrated Maritime Strategy, a step toward a holis- tic policy on this matter. Maritime security is also an integral part of the IMO's responsibilities. In recent years, Africa has witnessed a resurgence of the problems of piracy, human trafficking, and dumping of toxic waste in its coastal waters--in addition to illegal fishing, which has been going on for decades unnoticed by poorly equipped African states. In particular, the rapid esca- lation of piracy off the coast of Somalia and the Gulf of Guinea has alarmed African states as well as the international community. A comprehensive security regime for international shipping entered into force on July 1, 2004. The mandatory security measures, adopted in December 2002, include a number of amendments to the 1974 Safety of Life at Sea Convention (SOLAS), the most far-reaching of which enshrines the new International Ship and Port Facility Security Code (ISPS Code),9 which contains detailed security-related requirements for governments, port authorities, and shipping compa- nies in a mandatory section (part A), together with a series of guidelines about how to meet these requirements in a second, nonmandatory sec- tion (part B). Most ports in Africa now have approval under the ISPS Code, but information on additional measures is sparse. A number of ports have closed-circuit television and automated port processes using electronic means to collect, analyze, and distribute data to improve real-time under- standing and control of the port situation. The frequency of container scanning has also increased as part of supply-chain security initiatives. Lax security is particularly evident among smaller secondary ports. On the 212 Africa's Transport Infrastructure other hand, oil export terminals and South Africa's seven commercial ports have particularly strong security. Meeting new industry security requirements (particularly those of the ISPS Code) can be expensive, especially for ports in developing countries. A recent study (Kruk and Donner 2008) examined compliance costs for 12 ports in both developed and developing countries. The average cost per TEU of container traffic was found to be $4.95, and the average cost per tonne of general cargo $0.22. The United Nations Conference on Trade and Development performed a similar study that gave substantially lower estimates, with average costs of $3.60 and $0.08, respectively. But Kruk and Donner (2008) found average security costs to vary widely among ports, ranging from $1.00 to $14.00 per TEU of container traffic and from less than $0.05 to $0.50 per metric tonne of general cargo. The three African ports had costs at the lower end of both ranges--between $1.00 and $2.00 per TEU and less than $0.05 per tonne of general cargo. Based on the results, the total cost of ISPS Code compliance for the three African ports taken together is just over $5 million. Performance, Cost, and Quality Measuring and comparing cargo-handling performance is feasible in certain cargo sectors but not really practical in others. For example, both container handling and general cargo handling are uniform enough to permit performance comparisons among ports, regions, and even continents. In measuring container-handling performance, both quay- side and landside performance share enough common denominators to be relevant. (Performance measures for the region can be found in appendix 5d.) Bulk handling, on the other hand, tends to be a bespoke business ranging across a wide variety of different commodities that use diverse handling systems. Performance comparisons in the dry bulk sec- tor are therefore generally uninformative. Quayside Container Handling: Below International Standards Quayside container-handling performance is measured by the average number of crane moves per hour. Figure 5.7 shows the performance of the major African ports. African ports perform poorly in this respect compared to other regions. Container-handling performance in modern container terminals utilizing container gantry cranes falls mainly in the 20­30 moves per hour Ports and Shipping: Moving toward Modern Management Structures 213 Figure 5.7 Container-Handling Systems at Major Ports of the Region Abidjana Dar es Salaama Douala concessioned Toamasinab Djibouti Tema Maputo Beira Apapa concessioned recently Dakarc port Onne Luandad Durban Elizabeth Cape Town not concessioned Mombasa East London Walvis Bay Port Sudan Pointe Noire Matadi 0 5 10 15 20 25 average moves per hour container gantries mobile cranes ships' gear Source: Mundy and Penfold 2009. Note: Handling rates reported are per hour, per item of equipment for container gantries and mobile cranes, and overall per hour for ships' cranes. a. Abidjan and Dar es Salaam average more than 20 moves per hour. b. Toamasina average moves for discharge (22.4) and moves for loading (12.8). c. Dakar only concessioned in 2007. d. Luanda only concessioned in 2007. 214 Africa's Transport Infrastructure bracket, while the norm for ship-to-shore handling performance in the African ports considered is below 20 moves per hour (table 5.10). Differences in handling volumes can explain part of the discrepancy in handling performance. Container terminals in Africa--particularly along the east and west coasts--handle much lower volumes than highly developed container terminals in other parts of the world, which translates into fewer exchanges per ship. Vessels in the region are also smaller, necessitating more complex vessel loading and unloading oper- ations. Moreover, a significant number of ports in Africa do not as yet possess purpose-built container-handling cranes and rely on ships' gear for across-the-quay handling operations. Eight moves per hour is the norm for this latter type of operation. Concessioning typically involves both investment in new equipment and improvement of management systems; the better management has also allowed ports to substantially enhance their handling performance without major upgrades of their quayside equipment. For example, within a few months of the concessioning of the Apapa container termi- nal in Lagos, delays for berthing space had dwindled, and leading shipping lines had reduced the congestion surcharge (charged to customers for moving traffic through the port) from $800 to $120 per container, saving the Nigerian economy an estimated $310 million per year. Similarly, the incoming operator in Dakar, DP World, has realized significant Table 5.10 Gantry Crane Productivity, 2004 (Selected Terminals) Performance in net container Port moves per crane-hour United States Virginia International Terminals 25.0­30.0 Ceres, Baltimore 24.0 Europe Antwerp (Scheldt) 28.0 Rotterdam Delta 27.0 Asia Port of Singapore 33.0 Port Klang (Northport) 22.6 Africa Durban 15.0 Cape Town 12.0 Mombassa 10.0 Abidjan 20.0+ Source: Mundy and Penfold 2009. Ports and Shipping: Moving toward Modern Management Structures 215 operational improvements and achieved better performance through new equipment provision and the introduction of modern container ter- minal operating practices. Figure 5.8 shows that concessioned ports tend to outperform nonconcessioned ports in Africa, handling an average of 16 container moves per hour on average, compared to only 10 for pub- licly managed ports. They also have a much lower probability of relying on ships' gear for loading and unloading. Landside Container Handling: A Problem of Organization There are two major indicators of landside container terminal perform- ance: truck cycle time and the average dwell time of containers in the ter- minal (figure 5.9). The truck cycle time measures the time between when a truck joins the queue to enter the terminal (to drop off containers) and when it exits the terminal (loaded with other containers). Based on international experience, the benchmark for truck cycle time is one hour, though the growth in container traffic has made achieving that bench- mark increasingly rare. The regional average for truck cycle times in Africa ranges from 4 hours in southern Africa to 10 hours in West Africa. Improving the unsatisfactory landside performance will require address- ing basic terminal organization; using prebooking and IT systems; and Figure 5.8 Container-Handling Performance by Equipment Type and Port Management Type 20 16 average moves per hour 15 14 10 10 8 5 0 container gantries ships' gear concessioned not concessioned and mobile cranes handling system concession status Source: Mundy and Penfold 2009. Note: Handling rates reported are per hour, per item of equipment for container gantries and mobile cranes, and overall per hour for ships' cranes. 216 Africa's Transport Infrastructure Figure 5.9 Typical Truck Cycle and Dwell Times by Region 35 30 30 28 25 24 24 number of times 20 15 15 12 11 12 10 10 8 6 5.5 6 5 3.5 4 4 5 2 0 minimum average maximum minimum average maximum truck cycle times (hours) dwell times (days) Southern Africa East Africa West Africa Source: Mundy and Penfold 2009. introducing off-peak pickup times, modern gate systems, and other meas- ures that alleviate congestion around the terminal (such as relocating stuffing and stripping activities to satellite sites). The accepted international target for container dwell time is less than seven days. Dwell times in Sub-Saharan Africa range from an average of 6 days in southern Africa to 12 days in East Africa and more than 15 days in West Africa (see figure 5.9). Southern Africa's superior performance can be attributed to better container storage organization. Unlike terminals in East and West Africa, most southern African terminals charge the consignee a daily storage charge after five to seven days, and charges sometimes increase the longer a container remains in storage. Furthermore, terminals typically have rules to discourage the dumping of empty containers in a terminal. (A range of access and landside quality indicators are shown for all ports in Africa in appendix 5e.) General Cargo Handling: Below International Performance Standards In developed countries, handling rates for general cargo usually exceed 30 metric tonnes per hour per crane. Only the South African ports of Richards Bay and Durban approach this level of performance (figure 5.10). More generally, ports in Africa fall far behind the developed Ports and Shipping: Moving toward Modern Management Structures 217 Figure 5.10 Performance in General Cargo Handling in Individual Ports of Sub-Saharan Africa Richards Bay Durban Luanda Southern Africa Port Elizabeth Cape Town East London Maputo Toamasina Majajanga Beira Dar es Salaam East Africa Mombasa Sudan Suakin Cotonou Tema Takoradi West Africa Apapa Onne Harcourt Calabar Mindelo Point Noire Central Africa Matadi Boma 0 5 10 15 20 25 30 general cargo crane or gang productivity average (tonnes per hour) Source: Ocean Shipping Consultants 2009. country standard and will require modern systems and cargo-handling practices to bridge the gap. While there is substantial variation in general cargo handling perform- ance within each subregion in Africa, there is also a clear difference in average performance between regions, with West Africa falling substan- tially behind East and southern Africa (table 5.11). 218 Africa's Transport Infrastructure Table 5.11 Performance in General Cargo Handling across Regions of Africa Performance Region per hour per crane East Africa 8­25 metric tonnes Southern Africa 10­25 metric tonnes West Africa 7­15 metric tonnes Source: Mundy and Penfold 2009. Dry and Liquid Bulk Cargo Handling: International Companies Meeting International Performance Standards Both dry and liquid bulk cargo comprise a wide range of commodities, which makes measuring and comparing performance in this sector diffi- cult. Moreover, privately owned industrial-style terminals that handle dry and liquid bulk cargo tend to keep their performance statistics confiden- tial. Nevertheless, some conclusions about performance in the sector can be drawn. The performance of private operators handling dry bulk and liquid bulk (particularly oil) cargo in Africa is likely to be comparable to that of similar terminals in mature markets because they are mostly global oper- ators well versed in state-of-the-art techniques. On the other hand, gen- eral cargo quays handle a considerable proportion of dry bulk cargo in the region, which suggests that the low traffic volume of certain bulk cargo types does not justify dedicated facilities. There is therefore scope for the specialization of facilities as volumes increase. Cargo-Handling Costs: High for All Cargo Cargo-handling costs tend to be higher in Africa than in mature mar- kets in most other parts of the world. This is a result of several factors. Technical deficiencies (including low operating efficiency, lack of maintenance, poor planning, and capacity constraints) and institutional deficiencies (including a lack of enterprise culture, outdated pricing structures, and weak regulation of the monopoly service provider) con- tribute to inflated costs. (Available data on costs and charges at African ports are given in appendix 5f.) Container-handling charges at ports in Africa range from $100 to $300 per container, compared to between $80 and $150 per container elsewhere. Shipping lines are frequently involved in terminal operations in the region, particularly along the coast of West Africa. Shipping line involvement in terminal operations, either direct or indirect, has raised questions about Ports and Shipping: Moving toward Modern Management Structures 219 possible abuse by shipping lines to gain market share in the ocean freight markets. Concessions should therefore include regulations to safeguard against this possibility. Such regulations might involve specifying that port installations are for common use and should not discriminate in the charges levied. Concession agreements did not place ceilings on container-handling charges until recently; ceilings were specified in the concessioning of container-handling facilities in Madagascar in 2005 and Nigeria since 2004. Handling charges for general cargo are between $6 and $15 per tonne in the region, compared to between $6 and $9 per tonne elsewhere. A lack of proper facilities and equipment and ineffective management and operations generally contribute to higher charges. Furthermore, typical breakbulk cargo10--unlike general cargo such as palletized fruit or bagged agricultural products--is generally handled by publicly owned operators, which can reduce efficiency and increase costs. Charges vary by trade lane and commodity. Figure 5.11 shows the range of basic handling charges from ship's hold to gate enforced by the port or terminal. These do not include the inevitably higher charges passed on by the shipping line to cargo shippers or supplemental charges such as for scanning or congestion. (More details on cargo-handling charges can be found in appendix 5f.) Overall Quality of Port Services: Substantial Improvement Needed With the possible exception of those in South Africa, ports in Africa are comparable to those in other emerging regions. If they are to catch up to ports in more developed areas of the world, they must make substantial improvements in port planning, infrastructure development, institutional reform, pricing structures, and interface with other transport systems. There are also external catalysts for improvement--for instance, the advance of global liner operators into the African region linking the con- tinent with global liner networks. Africa is no longer served only by regional specialists. This has spurred port traffic growth and international investment interest in the ports sector. Finally, governments also must play a role in encouraging reform. Overall, the context is conducive for improving the quality of port services. The Way Forward Growth in maritime traffic in Africa promises to continue--as does pressure on the region's port system--once the effects of the financial crisis of 2008­09 have worn off. The liberalization of the international Figure 5.11 Typical Gateway Container-Handling and General Cargo Charges Applied in Major African and World Ports 220 350 320 15 15 15 general cargo (over-the-quay per metric ton, US$) 300 15 275 container handling (ship to gate, US$) 250 243 11 200 9 9 10 8.5 8 7.5 8 154 135 150 7 7 130 144 110 110 6 96 100 100 100 95 5 80 50 0 0 sia pe ia m pe sia t ica a a a a ca ca as ric ric ric ric As do fri fri ro ro tA la er rE Af Af Af Af tA tA ra h Eu ng Eu Am Fa as ut rn rn st st st es es Ki he rn n Ea Ea Au So he he tin er W W d he ut ut ut rth ite d La So ut an So So Un No So st Ea e rest of the world Africa dl id ports M rest of the world Africa minimum Africa maximum general cargo Source: Mundy and Penfold 2009. Ports and Shipping: Moving toward Modern Management Structures 221 shipping market will continue to have a positive and widespread influ- ence; neither the quality nor quantity of international shipping is likely to be deficient. The region's ports, however, have not kept up with the increase in traffic or changes in technology and regulation. Their performance lags behind that of most ports around the world. Several ports have exceeded their capacity limits, and others are approaching the limit. The presence of international container liner operators has stimulated port development in the region, but slow institutional and regulatory reform is a bottleneck to progress. Further concerns include corruption and unsatisfactory integration between maritime and land transport. The priorities for action therefore are primarily aimed at strategic ele- ments and institutional reform. Priority 1. Improving strategic planning for ports Governments can improve strategic planning through the following: · Establishing a strategic port development plan · Coordinating planning for port facilities and local land use · Seeking international agreements on port locations and developments, and on international transit corridor management. Priority 2. Reforming port management structures Most countries have replaced the traditional public sector service port structure, but many African countries have failed to do so. Reform should therefore include the following: · A formal commitment to a modern landlord port structure or its equivalent · Procedures for planning and implementing tenders for port concessions · Transparent rules and procedures (including for international arbitra- tion) to attract the best operators to concessions. Priority 3. Reforming public administration Customs administration is inherently susceptible to corruption and inef- ficiency. Public administration in the region requires reforms such as these to minimize opportunities for manipulation: · Automating customs procedures and other procedures, such as health and safety inspection 222 Africa's Transport Infrastructure · Developing portwide multiuser hardware and software systems to speed processing · Establishing an independent port regulator. Priority 4. Improving coordination with land transport An efficient port minimizes the cost and time of transfers between land and sea transport. Improving this aspect of ports in the region will require taking these steps: · Eliminating non-market-based allocation of inbound traffic to land transport operators (including both the protection of preferred carri- ers and noncompetitive procedures for the allocation of traffics) · Providing adequate infrastructure for road and rail access to port areas. Notes 1. The main source for this chapter is a report prepared for Africa Infrastructure Country Diagnostic in 2008 and 2009 by Ocean Shipping Consultants, Ltd. (2009). The report covers traffic development, infrastructure development and investment, performance cost and quality, institutional and regulatory frameworks, and security arrangements. Bert Kruk and Michel Donner of the World Bank also contributed extensively to this chapter through their exam- ination of the performance of shipping markets in Africa. 2. Containers with a standard cross section of 8 feet by 8 feet may be 10, 20, or 40 feet in length. For statistical purposes, it is conventional to express con- tainer traffic in TEUs. 3. DP World, a subsidiary of the investment company Dubai World, is a major operator of marine ports, with 49 terminals in operation (and a further 12 under development) that handled 46.8 million TEUs in 2008. 4. In 2004, the president of Côte d'Ivoire announced the award of a concession to operate the Vridi container terminal to a company (SETV) largely owned by the French logistics company Bolloré. The award was made without any competitive tendering. It was challenged by the chamber of commerce, other potential concessionaires, and also by the minister of infrastructure, who declared the award was illegal as it was sanctioned by neither his ministry nor the government. The minister of infrastructure was subsequently suspended by the president. Following the refusal by the Ministry of Transport to renew Bolloré's stevedoring license, the company took the case to the highest court in Côte d'Ivoire, which overruled the Ministry of Transport's decision. The World Bank withdrew its support but the concession went ahead. A compro- mise eventually emerged, with APM Terminals invited to hold a 47 percent Ports and Shipping: Moving toward Modern Management Structures 223 stake in the container terminal concession. Container movements reached 544,000 TEUs in 2008, and SETV has pledged to invest $62 million to fur- ther expand the capacity of the terminal. 5. The United Nations Liner code, which aims to protect the interests of devel- oping countries by reserving 40 percent of the trade of liner conferences for ships of the country of origin or traffic destination, was introduced in 1974. But the container revolution largely undermined its effectiveness by moving a large proportion of the international freight movement out of the confer- ence system, and attempts to adapt it to the new world container system have been unsuccessful. 6. ASYCUDA is a computerized customs management system, developed by the United Nations Conference on Trade and Development in Geneva, that covers most foreign trade procedures. The system handles manifests and cus- toms declarations, accounting procedures, and transit and suspense proce- dures, as well as generating trade data for statistical economic analysis. It provides for electronic data interchange between traders and customs using EDIFACT (Electronic Data Interchange for Administration, Commerce, and Transport) rules. ASYCUDA maintains permanent regional support centers in Ouagadougou, Burkina Faso, for West Africa, and in Lusaka, Zambia, for East and southern Africa. 7. A quay is a concrete, stone, or metal platform lying alongside or projecting into the water for loading and unloading ships. A container gantry crane is a track-mounted, shoreside crane used in the loading and unloading of break- bulk cargo, containers, and heavy-lift cargo. 8. A ship carrying a flag of convenience is a ship registered under the maritime laws of a country that is not the home country of the ship's owners and that offers low tax rates and leniency in crew and safety requirements. 9. The code is a comprehensive set of measures established in the wake of the 9/11 attacks in the United States. Compliance is mandatory for the 148 signa- tories of the SOLAS. 10. Breakbulk cargo covers a great variety of goods that must be loaded individ- ually and not in intermodal containers or in bulk. References Bell, M. G. H., and K. Bichou. 2007. The Port Sector in South Africa: Towards an Integrated Policy and Institutional Reform. Washington, DC: World Bank. CICOS (Commission Internationale du Bassin Congo-Oubangui-Sangha). 2007. "Plan d'action stratégique pour la promotion de la navigation dans le bassin Congo-Oubangui-Sangha." HPC Hamburg Port Consulting GmbH, Hamburg, Germany. 224 Africa's Transport Infrastructure de Wulf, L., and J. B. Sokol, eds. 2004. Customs Modernization Studies. Washington, DC: World Bank. ------. 2005. Customs Modernization Handbook. Washington, DC: World Bank. Drewry Shipping Consultants. 2006. Annual Review of Global Container Terminal Operators. London: Drewry Shipping Consultants. ------. 2009. Annual Review of Global Container Terminal Operators. London: Drewry Shipping Consultants. Harding, A., G. Palsson, and G. Raballand. 2007. "Port and Maritime Transport Challenges in West and Central Africa." Sub-Saharan Africa Transport Policy Program Working Paper 84, World Bank, Washington, DC. Kruk, C. B., and M. L. Donner. 2008. "Review of Cost of Compliance with the New International Freight Transport Security Requirements: Consolidated Report of the Investigations Carried Out in Ports in the Africa, Europe and Central Asia, and Latin America and Caribbean Regions." Transport Sector Board, World Bank, Washington, DC. McTiernan, A. 2006. "Customs and Business in Africa: A Better Way Forward Together." Business Action for Improving Customs Administration in Africa, London. Mundy, M., and A. Penfold. 2009. "Beyond the Bottlenecks: Ports in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 13, World Bank, Washington, DC. Ocean Shipping Consultants. 2009. "Beyond the Bottlenecks: Ports in Africa." Africa Infrastructure Country Diagnostic Background Paper 8, World Bank, Washington, DC. U.S. Department of Energy. 2006. "Annual Energy Review." Energy Information Administration, U.S. Department of Energy, Washington, DC. World Bank. 2007. Port Reform Toolkit. 2nd ed. Washington, DC: World Bank. ------. 2008. Private Participation in Infrastructure Database. World Bank, Washington, DC. http://ppi.worldbank.org/index.aspx. CHAPTER 6 Urban Transport: Struggling with Growth Urban transport is not a mode of transport, of course, but rather a collection of modal facilities and services found in a particular location.1 It is the density and complexity of those facilities and services that differenti- ate one system of urban transport from another. This chapter, unlike those that precede it, focuses on services more than physical infrastructure. For the infrastructure facilities discussed, data are not comprehensive but are based on a sample of 14 cities.2 The sample data are supplemented from other sources where they are clearly unrepresentative (as in the case of urban rail services). All of the sample cities are in low-income countries (table 6.1). The sample includes medium cities (population around 1 million), large cities (population of 2­ 4 million), and the two megacities (population over 5 million) of Lagos and Kinshasa. Density also varies across the sample, from under 1,000 inhabitants per square kilometer (km2) in Kigali and Kinshasa to over 13,000 in Conakry and Douala. For comparison, the density of New York City is around 10,000 inhabitants per km2. Africa's cities are experiencing rapid population growth--typically between 3 and 5 percent per year over the past decade. Douala and Lagos each have grown by at least 6 percent per year. In 2000, one in three Africans lived in a city, and this share is expected to rise to one in two by 225 226 Table 6.1 Size and Other Characteristics of 14 African Cities City population as City population as Density Private cars Population Growth share of national share of national (1,000 inhabitants per 1,000 City (millions) rate (%) population (%) urban population (%) per square km) inhabitants Abidjan 3.5 3.7 20 46 6.2 52 Accra 2.8 4.0 13 28 8.1 65 Addis Ababa 3.1 < 4.0 5 28 5.7 32 Bamako 1.2 4.8 9 32 4.5 108 Conakry 1.5 4.1 16 65 13.3 61 Dakar 2.8 4.4 30 49 5.1 39 Dar es Salaam 3.7 6.4 10 29 2.1 13 Douala 2.5 6.5 17 34 13.2 40 Kampala 2.0 5.0 7 59 2.1 -- Kigali 0.7 3.1 8 100 1.0 40 Kinshasa 8.0 4.1 13 32 0.8 -- Lagos 15.0 6.0 12 25 4.2 80 Nairobi 4.0 > 4.0 12 29 5.7 -- Ouagadougou 1.1 4.4 8 52 2.0 78 Average n.a. 4.7 13 43 5.2 55 Source: Kumar and Barrett 2008. Note: n.a. = not applicable; -- = not available. Urban Transport: Struggling with Growth 227 2030. The growth has been driven by anemic economic conditions in rural areas rather than by burgeoning wealth in the cities, with people fleeing rural areas to escape failing crops, natural disasters, and conflicts. The challenge posed by rapid growth is accentuated by the absence of policies on land use and economic development. This has led to urban sprawl, as migrants from rural areas settle in outer areas where land is most cheaply available. The declining population density associated with this sprawl has increased travel distances and pushed up the price of public transport. In most cities, authorities have had difficulty meeting the serv- ice demands of the new urban residents, particularly the poor, who are most dependent on the public provision of water, electricity, transport, and other services. As a result, the poor are often effectively excluded from work and social services. Meanwhile, the rising use of private cars has choked roads, endangering the safety of pedestrians and the health of city residents who breathe in automobile emissions. The need to coordinate land use and transport planning is widely recognized but presents a very difficult challenge to urban governance. Infrastructure: Roads Urban transport in African cities is largely road based. Most roads were built when the cities had a single center and before the rapid growth of personal motorized transport. The characteristics of the road networks in the 14 study cities are summarized in table 6.2. The primary road net- work usually radiates from the center of the city to surrounding areas but lacks orbital links. As a consequence, the region's urban road infrastruc- ture is deficient in several ways. First, there is not enough of it. Many African cities have expanded more quickly than the capacity of governments to provide infrastructure. As a result, road networks are incomplete and unconnected. New hous- ing construction has been largely unplanned, without adequate provision for transport and other services. Overall, the road network constitutes less than 7 percent of the land area in most of the 14 study cities--only about one-third that in most developed cities worldwide (World Bank 2002). Service lanes are absent, and street lighting is minimal. The majority of the roads have one lane in each direction; where the roads are wider, one lane is often taken up by pedestrians and parked vehicles. It is also diffi- cult to organize public transport services to serve areas of low density, and many outlying neighborhoods can be reached only by two-wheeled vehicles. 228 Africa's Transport Infrastructure Table 6.2 Characteristics of City Road Networks Length Length of Paved roads Paved road density of road paved road as share of m per km per City network (km) network (km) all roads (%) 1,000 people km2 Abidjan 2,042 1,205 59 346 2.1 Accra 1,899 950 50 339 2.8 Addis Ababa -- 400 -- 129 0.7 Bamako 836 201 24 167 0.8 Conakry 815 261 32 174 2.3 Dakar -- 1156 -- 467 2.1 Dar es Salaam 1,140 445 39 122 0.2 Douala 1,800 450 25 237 2.4 Kampala 610 451 74 225 0.5 Kigali 984 118 12 170 0.2 Kinshasa 5,000 500 10 63 0.1 Lagos -- 6,000 -- 400 1.7 Nairobi -- -- -- -- -- Ouagadougou 1,827 201 11 185 0.4 Average -- -- 33 318 1.7 Source: Kumar and Barrett 2008. Note: -- = not available. km = kilometer; m = meter. Second, the proportion of paved roads is too low. On average, only a third of the roads in the sample cities are paved (table 6.2) but the range is wide: from barely 10 percent in Kinshasa and Kigali to more than 70 percent in Kampala. Paved road density is typically in the order of 300 meters per 1,000 inhabitants (or close to 2 kilometers [km] per km2). These densities are at the extreme lower end of developing cities worldwide, for which the average is close to 1,000 meters per 1,000 inhabitants, according to the Millennium Cities database (UITP 2001). Again, the range is wide. Dakar has 467 meters of paved roads per 1,000 inhabitants; at the other extreme, Kinshasa has just 63 meters, barely half that of the city (Dar es Salaam) in second place for lowest paved road density. In low-income areas, gravel and earth roads are still the norm, and poor drainage contributes to serious flooding during the rainy season. Third, road infrastructure is poorly maintained. Although figures on road conditions are not widely available, the proportion in a poor state of repair is probably greater for urban roads of any specified daily traffic level than for nonurban roads with the same traffic volume. Estimates of the urban road stock were made for 20 of the 24 Africa Infrastructure Country Diagnostic (AICD) countries. The total network length was 268,490 km including South Africa, but only 104,250 excluding it. Urban Transport: Struggling with Growth 229 Outside South Africa, less than half of urban roads are paved, and only a little over 40 percent are in good condition (figure 6.1). The poor state of roads has also limited options for urban transport and contributed to the wear and tear of transit fleets. In their current state, many roads cannot handle a conventional bus service. Most of the countries surveyed have established a second-generation road fund (as discussed in chapter 2 of this book) and have begun to fund road maintenance through road use charges. Yet although most road use occurs in urban areas, and most of the revenues are therefore collected Figure 6.1 Condition of Urban Roads in 20 AICD Countries a. Including South Africa 60 50 40 percent 30 20 10 0 good fair poor condition b. Excluding South Africa 60 50 40 percent 30 20 10 0 good fair poor condition earth gravel paved Source: Gwilliam and others 2009. Note: Totals do not add up to 100 percent because of rounding. 230 Africa's Transport Infrastructure from urban road users, the fiscal allocation usually does not reflect this situation. In Ghana and Ethiopia, between 20 and 30 percent of the road fund is allocated to the maintenance of urban roads. On average across all countries, however, urban roads receive only about 10 percent of road fund revenues--an amount inadequate for their maintenance needs. While the implication is that cities are expected to have other sources of income from taxes or trading revenue that they can devote to their roads, they often do not. A restructuring of road fund allocations to more closely reflect traffic patterns could help to alleviate this problem. In addition, new sources of revenue need to be developed. For example, cities could charge private cars for parking and tax new urban develop- ments that impose a burden on existing transport networks. Fourth, urban roads are poorly managed. Intersections are spaced close together and are poorly designed for turning. In all the cities under review, commercial activities (such as street vending) and parked vehicles force pedestrians off the sidewalks into the roadway, reducing road capac- ity and posing safety hazards. Competition between vehicles for space at bus stops often spills over into adjoining traffic lanes. Because traffic man- agement is limited in scope and extent, accidents are frequent. Pedestrians account for two-thirds of traffic fatalities. Fifth, local governments have paid little attention to facilitating the operation of public transport systems, which account for the majority of trips in all of the study cities. For example, dedicated bus lanes that speed the flow of public transport are rare; bus stops, bus shelters, and other facilities for passengers are scarce and in poor condition; bus bays are too narrow to accommodate multiple buses, so waiting buses often obstruct one lane of the road; and bus terminals, often in the heart of the city, are overcrowded and lack facilities for passengers. A few cities have intro- duced measures to improve bus travel, but they have not been properly enforced anywhere. Finally, most cities have ignored the needs of pedestrians. Around 65 percent of the road network lacks sidewalks, and those that do exist are poorly maintained, have open drains, and are susceptible to takeover by the expansion of adjoining properties. Pedestrian crosswalks and bridges are not found outside of the city center, where they lack signals and are rarely respected by motorists or enforced by the police. Pedestrians often jump median strips and road dividers in high-traffic areas, triggering fre- quent serious accidents. Facilities for bicycles and other forms of nonmo- torized transport are equally scarce, and the few bicycles in Kampala and Nairobi compete dangerously with motorized vehicles for road space. Urban Transport: Struggling with Growth 231 Infrastructure: Rails South Africa has by far the largest commuter rail networks in Africa. In Pretoria, Johannesburg, Cape Town, and Durban, Metrorail operates exten- sive EMU services,3 each of which carry around half a million or more commuters each day, and it runs much smaller loco-hauled operations in Port Elisabeth and East London. In total, it carries over 500 million paying passengers each year. Metrorail operated as a distinct business unit within Transnet until it became part of the South African Rail Commuter Corporation in 2006. It has a fleet of 4,200 carriages (about 70 percent of which are operational) and runs services over more than 2,000 route-km, some of which it owns and some of which belong to Spoornet. Also in South Africa, a concession has been awarded for a standard-gauge rapid (160 km per hour) regional system between Johannesburg and Pretoria. Construction began in 2006 and the first stage is due for completion in 2010­11. This system differs from other urban railways in Africa because it is a "middle class" line, primarily aimed at diverting traffic from private cars in order to reduce road congestion. Outside of South Africa, the Petit Train Bleu in Dakar has been for many years the only regular commuter service operating in the region. Since 1988, it has operated between Dakar and Rufisque on the main line of what is now the Transrail concession. Service is relatively frequent, with 19 pairs of trains reportedly carrying 25,000 passengers per working day. The Petit Train Bleu is operated by the Agence Nationale de Nouveaux Chemins de Fer, the agency responsible for the non-Transrail network in Senegal. In other African cities, commuter services have been on a small scale, generally comprising one or two loco-hauled return services per day-- one into the city in the morning and a return in the evening. Examples include Nairobi (on three routes), Lagos (one route), Accra (two routes), Harare (two routes), Bulawayo, Luanda (one route with six return serv- ices daily), and Maputo and Kinshasa (one route each). Annual passenger traffic is typically 1 or 2 million at most (Bullock 2009). Sporadic attempts to develop commuter services in some other cities (such as the Njanji service in Lusaka) have generally failed. There are pos- itive recent signs, however: a new service was inaugurated in Kaduna in 2008, and Accra has ordered new diesel multiple units for its suburban service. Several other cities also have plans to introduce modern com- muter networks. Lagos is planning to complete a network of seven lines, totaling 246 km, by 2025; work on the first two lines, totaling 64 km, is 232 Africa's Transport Infrastructure already under way (Mobereola 2008). These services, however, will almost certainly require substantial external funding to cover both capital investment and recurrent operating costs. To best stimulate commuter services, new transport authorities should be established separate from the existing railway authority, as in South Africa. Such an enabling framework is essential if urban rail systems are to meet the transport demand of Africa's megacities. Institutions Jurisdiction over urban transport in Africa typically is spread over mul- tiple tiers of government, an arrangement that inhibits policy integra- tion and administrative consistency. Central governments still dominate urban transport, although some functions have been devolved to local governments in several cities (see table 6.3). Unfortunately, local authorities often lack the institutional and financial capacity to execute these functions. Poor accountability and a lack of coordination among institutions have also hindered the development of effective transport strategies. The institutional arrangements for urban roads are particularly com- plex. Typically, several national and local bodies share jurisdiction, and separate legislation governs roads and transport services. In Conakry, for example, several institutions have responsibility for different segments of the road network. And in Ghana, responsibility for urban transport has devolved from central to local governments, which have neither the resources nor the technical expertise to carry out the functions assigned to them. Therefore, the Ministry of Transportation (through the Department of Urban Roads) is effectively responsible for road mainte- nance and development. Impediments to Integrated Policy Patterns of urbanization and land use drive demand for transport services and shape the context for road construction and passenger service provi- sion. Effective urban public transportation therefore requires integrated management of urban planning, infrastructure maintenance, and services. Yet in the 14 study cities, the same institution rarely houses all three of these functions. Even when responsibility is retained by the central gov- ernment, it is usually spread among several ministries. The authorities in most of the 14 cities under review lack institutional capacity in land use planning and transport, and there is no effective Table 6.3 Institutions with Responsibility for Public Transport in Cities Entity responsible for Independent Entity responsible for Entity responsible for building and maintaining transport City urban planning transport planning of urban roads authority Abidjan Central government AGETU (central) Agence des Routes (for AGETU central services) and municipalities (for feeders) Accra Town and Country Planning Department; Ministry of Transportation (central) Ministry of Transportation None Ministry of Local Government, Rural Development and Environment Addis Ababa Ministry of Works and Urban Ministry of Transport and Addis Ababa City Roads None Development (central) Communications (central) Authority Bamako DRCTU (municipal) Direction Nationale des Transports Direction Nationale des None Terrestres et Fluviaux (national) Routes (District of Bamako) Conakry National Land Transport Directorate Ministry of Transportation Central government (trunk); None city government (feeders) Dakar n.a. CETUD (central) CETUD CETUD Dar es Salaam Ministry of Land, Housing and Settlement Ministry of Infrastructure Development; Ministry of Infrastructure SUMATRA Development (central) urban councils; SUMATRA; Dar es Development; Tanzania Salaam Rapid Transit Roads; municipalities Douala Urban Community of Douala Ministry of Transportation Ministry of Works; Urban None Community of Douala; city governments (continued) 233 234 Table 6.3 Institutions with Responsibility for Public Transport in Cities (continued) Entity responsible for Independent Entity responsible for Entity responsible for building and maintaining transport City urban planning transport planning of urban roads authority Kampala n.a. n.a. UTODA None Kigali Ministry of Infrastructure (city of Kigali); Ibid. Ministry of Infrastructure None ONATRACOM (district) Kinshasa Ministry of Planning Ministry of Transportation Ministry of Public Works and None Infrastructure (municipal) Lagos Ministry of Local Government LAMATA Ministry of Public Works LAMATA Nairobi Local government Ministry of Transport Ministry of Public Works and None local governments Ouagadougou Ministry of Transportation (central); Ministry of Infrastructure Ministry of Infrastructure None commune of Ouagadougou Source: Kumar and Barrett 2008. Note: For details on each city, see appendix 6a. n.a. = not applicable; AGETU = Agence de Gestion des Transports Urbains; CETUD = Conseil Exécutif des Transports Urbains de Dakar; DRCTU = Direction de la Régulation et du Contrôle du Transport Urbain; LAMATA = Lagos Metropolitan Area Transport Authority; ONATRACOM = Office National de Transport en Commun; SUMATRA = Surface and Marine Transport Regulatory Authority; UTODA = Uganda Taxi Operators and Drivers Association. Urban Transport: Struggling with Growth 235 forum for communication between those responsible for the two func- tions. Land use is generally ad hoc, driven by developers' interests and informal settlements, and the transport system is expected to respond accordingly. Poor planning and communication, coupled with a lack of funding to address capacity constraints, has resulted in high levels of congestion. A consequence of the functional separation is that the development of road infrastructure has generally focused on improving the flow of (mostly private) vehicles rather than of people. Public transport repre- sents a significant share of traffic, so its passengers have benefited from road development. But public transport priorities within the road net- work have received little attention. For example, there has been little effort toward creating dedicated infrastructure for bus transit (see box 6.1). Some measures favoring bus travel, such as bus priority lanes, have been introduced, but they have not been properly enforced. Box 6.1 Introducing Dedicated Infrastructure for Bus Transit Over the past few years, the World Bank has supported government initiatives in several cities (Accra, Dar es Salaam, Lagos) to implement bus rapid transit (BRT) systems. These systems deliver fast, comfortable, and affordable mass transit. Exclusive right-of-way lanes allow BRT to emulate the performance and amenities of a modern rail-based transit system at a fraction of the cost. The BRT systems in Accra and Dar es Salaam are expected to begin operation by 2011. The BRT system in Lagos, which was launched in March 2008, adapted best practices in Bogota, Columbia, and Curitiba, Brazil, to the African context. The BRT buses operate on an exclusive lane along the curb, which runs through the city for a distance of 20 km. Route franchising, improved ticketing systems, improved bus quality, and infrastructure investment have also benefited the city's public transport. Lagos has already seen the positive effects of the initiative. Preliminary studies suggest that over 200,000 commuters use the BRT system daily. Compared to previous bus systems in Lagos, average fares are 30 percent lower, travel times are cut in half, and average waiting times are 35 percent shorter. Source: A. Kumar, private correspondence. 236 Africa's Transport Infrastructure The Strategy Vacuum International experience suggests that if urban transport development is to be successful, it must be guided by cohesive policy. Yet none of the 14 cities studied has clearly articulated an urban passenger-transport policy and made it available to the public. Instead, it appears that ad hoc policy decisions respond to political pressures as they arise. This approach reflects the lack of a capable authority to coordinate regulation, transport planning, and infrastructure development. Such an authority requires a sufficient budget to attract and retain qualified staff and cannot be sub- ject to arbitrary changes when the authority in charge of transport comes under pressure. Reasonable user charges, such as fees for operating per- mits or franchises levied on operators, could provide needed funding. In practice, only a few of the 14 African cities have established agen- cies with overarching responsibility for urban transport. Addis Ababa has a city transport authority, but it is not autonomous. The agencies in Abidjan (Agence de Gestion des Transports Urbains, AGETU), Bamako (Direction de la Régulation et du Contrôle du Transport Urbain, DRCTU), and Dakar (Conseil Exécutif des Transports Urbains de Dakar, CETUD) lack sufficient authority to implement their plans and must instead work through other agencies of government. In 2001, Dar es Salaam established the Surface and Marine Transport Regulatory Authority (SUMATRA), a multisector agency with regulatory authority over rail, road, and maritime transport services, but without broader responsibility for urban transport planning. Only the Lagos Metropolitan Area Transport Authority (LAMATA) has any responsibility for road infrastructure development beyond initial planning. Services A breakdown of modal shares for urban transportation services in the 14 cities is presented in table 6.4.4 Not all of the cities have performed sufficiently detailed traffic sampling to produce reliable statistics, and oth- ers have limited their analyses to motorized transportation and excluded pedestrians. Nevertheless, some consistent transportation patterns can be found across the cities. As noted, urban public transport is largely road based. Buses (includ- ing large buses and minibuses) are the most common mode of public transit in most cities, with a modal share ranging from more than 70 per- cent in Dakar and Kigali to just over 10 percent in Bamako and 8 percent in Ouagadougou. Except in Addis Ababa and Ouagadougou, minibuses of Urban Transport: Struggling with Growth 237 Table 6.4 Modal Shares of Transport in Cities (% of Trips) City Large bus Minibus Taxi Motorcycle Private car Walking Other Abidjan 11 19 29 0 18 22 1 Accra 10 52 9 0 13 12 4 Addis Ababa 35 20 5 0 7 30 3 Bamako 1 10 5 56 19 -- 9 Conakry 1 14 6 0 1 78 0 Dakar 3 73 6 6 11 -- 1 Dar es Salaam 0 61 1 1 10 26 1 Douala 10 0 13 12 2 60 3 Kampala 0 41 -- 20 35 -- 4 Kigali 1 75 10 0 10 5 0 Kinshasa -- -- -- -- -- High -- Lagos 10 75 5 5 5 High 0 Nairobi 7 29 15 2 -- 47 0 Ouagadougou 8 0 -- 58 14 -- 20 Average 7 30 8 12 12 37 4 Source: Kumar and Barrett 2008. Note: -- = not available. Rows may not total 100 percent because of rounding. Percentages for Lagos exclude walking trips. up to 30-seat capacity are much more prevalent than large buses. In many cities, they have colloquial names--such as "tro-tro" in Accra, "danfo" in Lagos, "gbaka" in Abidjan, "sotrama" in Bamako, "matatu" in Nairobi, and "dala-dala" in Dar es Salaam--often relating to their origin or character- istics.5 Midibuses are larger than minibuses, with passenger capacities of between 30 and 50 (including standing room). The passenger capacity of a bus of a given size varies depending on the load limits and seating rules of each jurisdiction. Midibuses also have colloquial names such as "cars rapides" in Dakar and "molue" in Lagos. Of the 14 cities, only Douala and Ouagadougou do not offer a minibus service. The government of Cameroon outlawed minibuses to stimulate development of a new large- bus operator, which explains the absence of services in Douala. In both cities, however, shared taxis fill the transportation void left by the lack of minibuses. Overall, minibuses have a modal share that is twice that of large buses. To a certain extent it appears that larger buses and minibuses serve different routes. Large buses are more common on longer suburban routes, where their size allows them to offer lower fares. By contrast, minibuses tend to dominate congested areas, where their relative maneuverability allows them to charge higher fares. 238 Africa's Transport Infrastructure In recent years, the poor state of roads and the inability of bus compa- nies to maintain supply has led to an increase in the use of motorcycles for commercial transport in Douala, Lagos, and Kampala. The superior maneuverability of motorcycles allows them to more easily avoid pot- holes and navigate broken surfaces. Initially, motorcycle services provided access from residential areas to main roads, where passengers would then take taxis or buses. Motorcycle services now serve main roads and even the city center. Most of the motorcycles used for urban public transport have small engines (less than 100 cc [cubic capacity]). Drivers are often young and inexperienced and are not required to have a license, which leads to frequent--and often fatal--accidents. Most drivers own their own motorcycles or buy them on lease, paying in installments until the vehicle is totally owned, a process that can normally be completed within a year. Despite their widespread presence, motorcycle taxis are not ubiqui- tous; for example, they are rare in Nairobi and almost unknown in Dakar. In comparison with Asian cities, African cities have few nonmotorized vehicles such as bicycles and rickshaws, probably due to the state of the roads. Not all cities have measured the prevalence of walking as a mode of transport, but for those that have, the figure varies enormously. For example, walking has a modal share of 60 to 80 percent in cities such as Conakry, and Douala. By comparison, estimates for Abidjan, Accra, and Addis Ababa range from 10 to 30 percent. Railways generally play an insignificant role in urban transport in Africa. Dakar, Kinshasa, Lagos, and Nairobi have small suburban rail net- works, but none has a modal share of more than 2 percent. Suburban rail systems are an important part of urban transport in major cities only in South Africa, the main new prospect being Gautrain, a regional high- speed line between Johannesburg and Pretoria. In the future, bus rapid transit (BRT) systems may offer a less expen- sive form of public transport on a track segregated from other road traf- fic. Lagos was the path breaker--it opened the first 22 km of a system in 2008. Compared with some of the well-known Latin American systems, it is cheap and simple. It cost only $1.7 million per kilometer to build, but is not yet integrated with other modes or served by park-and-ride or feeder routes. It does not have electronic ticketing, but it is cheaper and quicker than other public transport alternatives and already carries 10 percent of the traffic to Lagos Island, with 25 percent of the travelers on its corridor in only 4 percent of the vehicles. Other countries are following suit. Johannesburg plans a 300-km sys- tem called Rea Vaya, with three integrated types of service (trunk, feeder, Urban Transport: Struggling with Growth 239 and complementary) that will be integrated with the high-speed rail link to Pretoria and to the airport. Smaller systems were developed for other South African cities that were 2010 World Cup venues. At the time of this writing, Dar es Salaam was planning a BRT system with a route length of 130 km, 18 terminals, and 228 stations, with the intention that construc- tion would commence toward the end of 2010 and take about two years to complete. Dakar and Accra also had schemes in development. Buses Large and Small Since the early 1990s, the 14 cities have taken different approaches to large-bus services (table 6.5). Accra, Dar es Salaam, Kampala, Kigali, and Lagos abandoned large-bus services altogether and now rely on private-- and largely informal--minibus services. Accra, Kampala, and Lagos have attempted to revive large-bus operations, so far without success. The government of Addis Ababa, on the other hand, continues to sub- sidize the city's public large-bus company (Anbessa), which it has allowed to remain in operation even though fares have been frozen since the early 1990s. In recent years, however, even Anbessa has faced a financial squeeze, and smaller buses have taken over a large share of the market. Other cities, primarily in the Francophone countries of West Africa, have adopted a range of private and public-private solutions, although some are too recent for their success to be measured. Abidjan and Ouagadougou established large-bus services that are run by private oper- ators but rely on government funding to cover operating deficits. Dakar and Douala have concessioned their large-bus services to private opera- tors, albeit with some government support. In Dakar, the large-bus com- pany is subsidized by the state. In Douala, one bus company has been given monopoly rights to operate on specified routes; other bus opera- tions are suppressed. Bamako, Conakry, and Kinshasa have fully privatized large-bus services that receive no government subsidies. Nairobi is the only city to have retained the private operation of its large- bus service since independence, although ownership has changed hands several times. The service was sold to the British operator Stagecoach in the early 1990s but was restructured as the Kenya Bus Service (KBS) in the late 1990s. It is owned and operated by the private sector. The minibus sector has flourished since the early 1990s, with fleet size growing at a rate of up to 11 percent per year in some cities. This growth increased imports of mainly secondhand vehicles accommodating from 15 to 30 passengers. Ownership of services is almost invariably informal; most individual owners have one or two vehicles that they rent out to 240 Africa's Transport Infrastructure Table 6.5 Large-Bus Operations in Cities Large-bus City operator Ownership Public support Abidjan SOTRA 60% public, 40% private Government covers deficits (subsidy per passenger for some categories). Accra MMT Ltd. 45% public, 55% private Government provides (including state-owned vehicle financing and enterprises) covers operating loss. Addis Ababa Anbessa 100% public (federal) City government provides subsidy per passenger. Bamako 8 different All private None operators Conakry Futur 100% private None Transporta Dakar DDD Private concessionaire Government provides operating subsidy, and donors help finance vehicles. Dar es Salaam None n.a. n.a. Douala SOCATUR Private concessionaire Service exclusivity exists for five years. Kampala None n.a. n.a. Kigali ONATRACOMb 100% public n.a. Kinshasa STUC 100% private None, but India helps finance vehicles. Lagos None n.a. n.a. Nairobi KBS 100% private None City Hoppa Ouagadougou SOTRACO 15% public Government provides subsidy (Ouagadougou in the form of exemptions commune), 85% private from taxes and duties. Source: Kumar and Barrett 2008. Note: n.a. = not applicable; DDD = Dakar Dem Dikk; KBS = Kenya Bus Service; MMT = Metro Mass Transit; ONATRACOM = Office National de Transport en Commun; SOCATUR = Société Camerounaise de Transports Urbains; SOTRA = Société de Transport Abidjanais; SOTRACO = Société de Transport en Commun; STUC = Société des Transports Urbains du Congo. a. Very few buses. b. Buses used mainly for intercity transportation. drivers. Drivers keep the fares they collect but are responsible for paying fuel costs, conductors' wages, terminal fees, and other incidental expenses. They therefore have a strong incentive to carry full passenger loads to maximize revenues and minimize variable costs (particularly fuel). Most minibus owners are government officials, businessmen, or professionals Urban Transport: Struggling with Growth 241 who seek to supplement their income while incurring minimal tax liabil- ity. Some owners, such as police and army officers and members of trans- port unions, can exploit their position to protect and enhance their businesses. An owner of a reasonably maintained vehicle with a reliable driver can generate sufficient revenues to pay for capital and operating costs, includ- ing adequate maintenance. On the other hand, owners often choose not to reinvest in the business when vehicles require major repairs. The sec- tor is therefore characterized by short ownership periods, few barriers to market entry or exit, and high turnover. In a few cities, formal minibus operators compete with informal oper- ators. Dakar, for example, has an estimated fleet of 3,000 cars rapides, 400 of which belong to one formal operator and 200 to another. In Dar es Salaam, a public bus company operates a fleet of 30 minibuses, which is a tiny share of the estimated 10,000 minibuses on the city's streets. And in Kinshasa, the private large-bus operator maintains 30 minibuses, a neg- ligible share of the estimated 1,200 minibuses that circulate the city. Across the study cities, minibus ownership is dispersed across many peo- ple, most of whom operate informally (table 6.6). Bus Fleets Large buses carry 50 to 100 passengers, although the upper end of that range includes standing passengers. Most large buses have a single deck and two axles, although a few double-decker buses are used in Accra. Articulated (three-axle) buses are found in only 1 of the 14 cities (Abidjan), and semiarticulated (four-axle) trailer buses have been used in Kinshasa. Most minibuses are light commercial vehicles converted to accommo- date passengers. Some were originally crew buses (buses operated by a driver and conductor). Almost all are integral-construction vehicles,6 although a few pickup conversions can still be found in Addis Ababa, where imports of integral small commercial passenger vehicles were for a time suppressed to protect the publicly owned operator (that policy has now been abandoned). In East Africa, the most popular vehicles are Japanese and are imported secondhand through traders in the Persian Gulf. Most minibuses in Nairobi and Kampala are diesel powered to economize on fuel costs, but the altitude and terrain of Addis Ababa make gasoline engines prefer- able. European vehicles are more popular in West Africa, although the region, particularly Abidjan, also has a significant number of Japanese 242 Table 6.6 Characteristics of Minibus Services in Cities Service designation Union or association (vehicle capacity in number for informal minibus City of passengers) operators Degree of formality of sector Ownership structure Abidjan Gbaka (22) UPETCA, Mainly informal Highly fragmented, 1­2 vehicles SNTMVCI per owner Accra Tro-tro GPRTU, PROTOA Mainly informal 80% of owners with 1 vehicle each Addis Ababa Minibus (8), midibus (22) Various Mostly informal 80% of owners with 1 vehicle each Bamako Sotrama, dourouni Yes Mainly informal Highly fragmented, 1­3 vehicles per owner Conakry Magbana (15­18) None Mainly informal Highly fragmented, 1­2 vehicles per owner Dakar Car rapide (23­32) Yes Two formal private companies, Two operators with large fleets, plus informal remainder highly fragmented Dar es Salaam Dala-dala (18­35) DARCOBOA Formal public company, Public operator with modest fleet, plus informal remainder highly fragmented Douala None (outlawed), shared taxis None Mainly informal Highly fragmented, 1­3 vehicles used instead per owner Kampala Known as taxis UTODA Mainly informal 80% of owners with 1 vehicle each Kigali Twegerane or shared taxi (14­20) ATRACO, Mainly informal Highly fragmented ONATRACOM Kinshasa Minibus (5­26) ACCO Formal private company, Formal operator with modest fleet, plus informal remainder highly fragmented Lagos Danfo (mini), molue (midi) Several affiliated with Mainly informal 80% of owners with 1 vehicle each NURTW Nairobi Matatu MVOA Mainly informal 80% of owners with 1 vehicle each Ouagadougou None, shared taxis used instead No Mainly informal Highly fragmented Source: Kumar and Barrett 2008. Note: ACCO = Association of Congo Chauffeurs; ATRACO = Association pour le Developpement de l'Artisanat du Rwanda; DARCOBOA = Dar es Salaam Commuter Bus Owners Association; GPRTU = Ghana Private Road Transport Union; MVOA = Matatu Vehicle Owners Association; NURTW = National Union of Road Transport Workers; ONATRACOM = Office National de Transport en Commun; PROTOA = Progressive Transport Owners Association; SNTMVCI = Syndicat National des Transporteurs et Voyageurs de Cote d'Ivoire; UPETCA = Union Patronale des Exploitants de Taxi-compteurs d'Abidjan; UTODA = Uganda Taxi Operators and Drivers Association. 243 244 Africa's Transport Infrastructure vehicles. In Lagos, the market came to be dominated by Volkswagen, whose vehicles have a flat area over the rear-mounted engine that pro- vides space for market goods. Many of the larger minibuses in Accra are Mercedes, and French vehicles are most common in Dakar. Most midibuses consist of a locally made body mounted on a light or medium truck chassis. Similar to the trend for minibuses, some Mercedes vans can be found in Accra, and Renault and Peugeot models are common in Dakar. Strong local assembly operations and dealer networks allowed Mercedes (specifically the 911 model) and Isuzu to dominate the markets in Lagos and Nairobi, respectively. Nairobi has also developed a niche market for luxury models in the more affluent suburbs. Many of the large buses operated in Africa were supplied new by donors and have an average age of 9 years. Unlike large buses, minibuses are typically purchased secondhand and therefore tend to be somewhat older, with an average age of 14 years. In East Africa, minibuses are typi- cally between 10 and 15 years old, although age varies widely throughout the region. Vehicles tend to be somewhat older in West Africa, and some larger minibuses are up to 20 years old (table 6.7). This partly reflects the Table 6.7 Average Bus Age and Fleet Size Large bus Minibus Average age Average age City (years) Fleet size (years) Fleet size Abidjan 7 650 15 5,000 Accra 1­2 600 15­20 6,000 Addis Ababa -- 350 -- 10,000 Bamako 17 168 15 1,800 Conakry 20 50 10­15 1,500 Dakar -- 410 15­20 3,000 Dar es Salaam n.a. 0 15 10,000 Douala 15 100 15­20 2,000 Kampala n.a. 0 10­15 7,000 Kigali 4 20 15 2000 Kinshasa 2 (STUC) 180 2 (STUC) 54 (STUC) 15­20 (informal) 1,200 (informal) Lagos -- < 100 > 15 75­120,000 Nairobi -- 250 > 15 10,000 Ouagadougou 5 55 n.a. 0 Average 9 218 14 11,400 Source: Kumar and Barrett 2008. Note: -- = not available; n.a. = not applicable; STUC = Société des Transports Urbains du Congo. Urban Transport: Struggling with Growth 245 region's less stringent technical enforcement. Body corrosion is the main reason for scrapping. Among cities in which large buses are still common, average age varies widely depending on access to subsidies and the influ- ence of public ownership. Midibuses can be even older than minibuses--some of the vans in Accra are approaching 30 years of age and others in Lagos are approach- ing 40 years. The government in Dakar is encouraging the replacement of its cars rapides, whose age can exceed 25 years. Midibuses tend to be more productive in Nairobi than elsewhere. Their replacement is there- fore more economical and they are generally not as old. Reliable data on vehicle productivity are scarce. Few buses have work- ing odometers, and drivers count the number of paying trips per day rather than the distance traveled. The limited evidence available indicates that both large buses and minibuses travel an average of 190 km per day (table 6.8). In some cities, however, the average distances traveled by large buses and minibuses differ widely. For example, minibuses in Abidjan, Addis Ababa, and Nairobi travel substantially farther each day than large buses. The opposite is true in Accra and Bamako. Where minibuses run low daily distances, it is usually because they are subject to the tour de role dispatching practices described in an earlier chapter. Table 6.8 Average Distance Traveled by Large Buses and Minibuses (kilometers per day) City Large bus Minibus Abidjan 161 250 Accra 160 140 Addis Ababa 138 180 Bamako 225 180 Conakry 180 180 Dakar 192 -- Dar es Salaam n.a. 180 Douala 180 n.a. Kampala n.a. 100 Kigali 210 210 Kinshasa 200 200 Lagos 180 100 Nairobi 200 240 Ouagadougou 250 n.a. Average 191 186 Source: City authorities. Note: -- = not available; n.a. = not applicable. 246 Africa's Transport Infrastructure Availability and Quality of Services Access to urban public transportation, as measured by seat availability per 1,000 urban residents, is much lower in Africa than in other regions. Most of the sample cities have 30­60 bus seats per 1,000 residents, although Addis Ababa, Kinshasa, and Ouagadougou each have no more than 10 per 1,000 (table 6.9). Overall, the 14 cities have an average of only 6 large-bus seats per 1,000 residents. For comparison, according to the World Bank's Urban Transport Indicators database, the middle- income countries of Latin America, Asia, the Middle East, and Eastern Europe (where private car availability is much higher) have an average of 30­40 large-bus seats per 1,000 urban residents. Based on the low density of paved roads, unplanned growth, poor road surfaces, and narrow streets in the region, it may safely be assumed that the geographic reach of bus services is seriously circumscribed in the 14 sample cities. Passengers suf- fer long waiting times, uncomfortable vehicles, and--potentially--the continued operation of unsafe vehicles. Formal and informal surveys of users undertaken in the 14 cities examined in Kumar and Barrett (2008) suggest widespread customer dissatisfaction with bus services. Frequent complaints include poor road quality, overcrowding of buses, unpredictable and irregular service, and Table 6.9 Availability of Public Transportation in Cities City Minibus Bus Taxi Total Seats per 1,000 population Abidjan 24 5 26 55 Accra 26 22 -- 48 Addis Ababa 4 6 2 12 Bamako 33 14 -- 47 Conakry 17 1 16 34 Dakar 27 7.5 15 48 Dar es Salaam 57 0 -- 57 Douala 16 4 27 47 Kampala 48 0 -- 48 Kigali 52 1 5 58 Kinshasa 4 2 -- 6 Lagos 60 0 -- 61 Nairobi 40 3 -- 43 Ouagadougou 0 1 6 7 Average 31 6 Source: Kumar and Barrett 2008. Note: -- = not available. Urban Transport: Struggling with Growth 247 inadequate terminal facilities. On average, passengers report walking for 10 minutes to reach a bus stop and waiting 30 minutes before a bus arrives. Trip times range from 30 to 45 minutes. Poor driving is prevalent in all 14 cities, and almost 50 percent of the operators interviewed iden- tified poor driver discipline to be a serious business problem. Bad behav- ior among drivers, especially in the vicinity of passenger pickup points and interchanges, contributes to low service quality. For example, compe- tition on common routes encourages drivers to block stops, drive aggres- sively, and stop in the roadway to pick up passengers, all of which increase traffic congestion. Governance of minibus operations does little to improve low service quality. Self-regulation by operators' unions, which is common, means that routes run between terminals are controlled by unions. This limits the ease with which routes can be adjusted to meet passenger demand, and many passengers must therefore change buses to reach their destina- tion, increasing the duration and cost of their trip. As a means of equitably distributing revenues and ensuring some discipline in operation, unions frequently require drivers to wait at terminals until their bus is fully loaded. As a result, passengers must walk to the terminal to secure a seat, then sit in it under a blazing hot sun to retain it. Waiting times at termi- nals can exceed one hour during off-peak periods. Finally, unions insist that vehicles be loaded in a strict rotation, which prevents passengers from rejecting vehicles that fail to meet expected standards for cleanliness or operating conditions. Under these circumstances, owners have little incentive to improve their vehicles, and investment in higher-quality vehicles is impractical. Three in four passengers interviewed in surveys for this study rated over- loading as their primary concern. Overcrowding of minibuses at the start of their journey is limited by the terminal management practices described above, as vehicles are dispatched once they are full. On the other hand, once the vehicles leave the terminal, there is little to stop drivers from over- loading, especially after dark. Overcrowding on large buses is more frequent and much worse. Performance data from Anbessa in Addis Ababa indicate that loads at peak times may be as high as 150 percent of rated capacity. During the morning peak time in Kinshasa and Dar es Salaam, load factors of large buses reach an average of 200 percent, and passengers are forced to hang out of the bus or sit on the roof. Nairobi attempted to mitigate over- crowding by outlawing standing passengers on KBS routes and restricting the number of passengers on matatu (minibuses). As a result, the comfort and safety of passengers improved, but fares also rose since operators had fewer customers to cover the same operating costs. 248 Africa's Transport Infrastructure Vehicle Maintenance Operators of large-bus fleets perform regular scheduled maintenance. Nevertheless, repairs are too frequent relative to maintenance, and even the younger fleets suffer from low levels of availability. For example, the availability of the core fleet in Addis Ababa was reported to be only 83 percent. Availability is also particularly low in Accra, although this is partly a result of a lack of technical support and spare parts for the fleet's new Chinese buses. Some small private operators in all 14 cities conduct basic preventive maintenance, but the intervals between oil and filter changes vary widely. More commonly, repairs are undertaken only when absolutely necessary to keep vehicles on the road. A consequence of poor maintenance is that the environmental per- formance of the minibus fleet has emerged as an important issue. For Lagos's large minibus fleet, the emissions problem is acute; in 2007, there were a few days when the city effectively closed down because of high pol- lution. Government agencies interviewed in several cities expressed some awareness of the problems associated with vehicle emissions, but only Accra (and to a lesser extent Dakar) has initiated a formal program to inform policy by quantifying the impact of emissions. There are two especially troubling sources of emissions: lead-based octane improvers in gasoline-powered vehicles, which continue to be used because of a lack of investment in refinery technology, and oil leaks caused by poor engine con- ditions. In the case of diesel engines, high sulfur levels in the fuel--again resulting from lack of investment in refinery technology--increase particu- late emissions, an effect compounded by poor maintenance of fuel-injection equipment. Most engines now in use predate the introduction of the first European standards in the early 1990s. Meeting the even higher standards of today would require new equipment and improved fuels. Most repairs, particularly of minibuses, are performed on the side of the road or in low-technology workshops using hand tools and no special- ized equipment. Lax and corrupt vehicle inspection regimes and low cap- ital investment in vehicles have allowed careless maintenance practices to remain sustainable. In some cities, obtaining a forged certificate of road- worthiness is reportedly easier and cheaper than passing an inspection test with a well-maintained vehicle. Cities are gradually recognizing the severity of the situation, with reforms initiated in Addis Ababa and planned in Kampala, where vehicle inspections are being privatized. But for the time being, many vehicle operators flout basic safety stan- dards for lighting, tires, and brakes. Routine vehicle inspections are clearly inadequate, and petty corruption among police officers prevents them Urban Transport: Struggling with Growth 249 from enforcing standards on the road. Overloading of vehicles is also a safety concern, although recent improvements in construction have allowed for increases in the permitted capacity of some vehicles. Costs: Fuel, Labor, and Taxes In the industrialized world, labor is the largest recurrent cost of conven- tional bus operation. But in African cities, fuel takes first place, account- ing for half of the total cost of operations. Fuel is also the largest cost for minibus operations. That cost has risen with international oil prices in recent years. While Accra and Lagos stand out for their relatively low fuel costs, diesel fuel and premium gasoline typically cost between $0.80 and $1.00 per liter--and even more in some cities (table 6.10). Because of the region's high unemployment and low wages, labor can account for less than 25 percent of total operating costs for large buses in Africa. The significance of this cost structure is that the viability of bus operations is even more sensitive to fuel costs in Africa than it is in industrialized countries. In all of the study cities, minibus operators also face nonoperational charges, such as petty extortion from enforcement agencies and local Table 6.10 Fuel Prices in Cities, July 2007 (US$ per liter) Premium City gasoline Diesel fuel Abidjan 1.25 1.09 Accra 0.49 0.43 Addis Ababa 0.60 0.42 Bamako 1.17 0.90 Conakry 1.50 0.69 Dakar 1.10 0.90 Dar es Salaam 0.93 0.87 Douala 0.95 0.83 Kampala 1.02 0.88 Kigali 1.15 0.99 Kinshasa 0.92 0.81 Lagos 0.71 0.83 Nairobi 0.92 0.76 Ouagadougou 1.18 0.94 Average 0.97 0.78 Source: City authorities; World Development Indicators (data assembled by Kumar and Barrett 2008). 250 Africa's Transport Infrastructure gangs, and payments to associations. The scale of these charges is difficult to assess accurately, and the claims of the operators are often strongly challenged by the associations themselves. Nevertheless, typical daily charges per minibus appear to range from about $1.50 in Accra to $10.00 in Kampala. Official association charges in Lagos are about $2 per day, and unofficial charges may be equally large. The associations reportedly do not declare the revenues raised through these charges to the tax authorities, but use them to enrich association officials and support grass- roots political interests. Taxing the minibus sector is made difficult by the fact that minibus drivers typically rent their vehicles from owners on a daily basis and do not formally record their fare revenues. Accra and Addis Ababa have both addressed this problem, at least in respect to vehicle owners. In those cities, authorities estimate the likely gross revenues and profit margins from owning various types of vehicles and assess income tax based on those figures. For example, in Addis Ababa the authorities estimate that the owner of a minibus taxi that is less than 15 years old will earn annual revenues of Br 25,000 ($2,900) and a profit of Br 6,620 ($770). The owner then must pay Br 482 ($56) in taxes on that profit. In Kenya, import duty on vehicles (135 percent of vehicle value, excluding value added tax) contributes to the high costs of matatu operation, and is also a cause of the high average age of the vehicles operated. Large-bus operators, by contrast, are subject to a range of business taxes, but investment incentives, such as accelerated depreciation allowances, usually provide the companies with some tax relief. Indirect taxes, such as duties on fuel and imported spare parts, can be very signif- icant. The large-bus operator in Nairobi once calculated that 24 percent of its costs were payments to the government. Nevertheless, efforts to relieve any one sector of the burden of indirect taxes are usually poorly targeted, with direct implications for the wider economy. For example, the government-mandated lowering of duties on spare parts for agricultural equipment in Kenya led tractor distributors to market common compo- nents to the wider transport sector. Conversely, any attempt to lower the burden on inputs in the bus sector would almost certainly leak to other sectors and to individuals. Fares Fare structures are very simple in Africa. Typically, a single flat fare applies to each route, with higher fares on longer routes. In cities where a single Urban Transport: Struggling with Growth 251 flat fare applies, route lengths tend to be determined in such a way as to avoid losses on longer routes. Setting and Controlling Fares Throughout the region, authorities are usually responsible for setting fares for services (table 6.11). But attempts to control large-bus fares are often unrealistic and counterproductive. In Addis Ababa, for example, fares for large buses officially remain at 1992 levels, and in Abidjan they have not been adjusted since 1994. Such artificially low fares have led to a drastic decline in the quality and coverage of large-bus services through- out the region. Only in Addis Ababa are public subsidies nearly large or regular enough to consistently cover operating deficits. Controlling fares in the informal sector has proved even more difficult. In West Africa, governments usually have formal control of fares for minibuses and shared taxis; this is the case in Bamako, Conakry, Dakar, and Douala, though in Ouagadougou an operating company, the Société de Transport en Commun (SOTRACO), determines all bus fares.7 Usually a set fare applies to passengers boarding at the departure termi- nal and along much of the route length, but passengers boarding close to the arrival terminal may be charged a lower fare. Passengers do not usu- ally receive a fare rebate for alighting before the arrival terminal, though this is not unheard of. In cities that also control minibus fares, authorities usually ensure that fares for large buses operated by public companies are set below minibus fares. In East Africa, in contrast, minibus operators have more flexibility to determine their own fares. In Dar es Salaam, the regulator SUMATRA allocates routes, and fares are subject to negotia- tions with the bus operators' association. In practice, attempts to impose price controls on the informal minibus sector usually fail to control real fares, but rather have the unintended consequence of route proliferation. In all 14 cities, minibus operators have shortened route lengths in response to fare controls, allowing them to charge fares that satisfy price controls but that increase trip costs for at least some passengers. Ultimately, both the cost of travel and travel times have increased. For instance, the official fare for a typical trip from Dakar to Pikine in Senegal is CFAF 110. But to make the complete trip, passen- gers must change vehicles twice and pay three fares totaling more than CFAF 200. Similarly, the fare for a trip of 5 km in Addis Ababa is set at Br 1.0, yet passengers routinely pay more than Br 1.5. Economic reforms in many of the cities have led to official deregulation of minibus fares, but in practice, some cities have retained partial administrative control. 252 Table 6.11 Fare-Setting Procedures for Large Buses and Minibuses Fares regulated for large Year of most recent Operating subsidy City buses and minibuses? Entity responsible for regulating fares fare adjustment for large-bus operators? Abidjan Yes AGETU 1994 Yes Accra Yes Ministry of Transportation 2004 Yes Addis Ababa Large buses only City government 1992 Yes Bamako Large buses only Operators/government 2006 No Conakry Yes Direction Nationale des Transports 2006 No Terrestres, operators' syndicates Dakar Yes Ministry of Finance 2000 Yes Dar es Salaam Yes Dar Commuter Bus Owners 2006 n.a. Association, SUMATRA, operators Douala n.a. Ministry of Finance n.a. Yes Kampala Yes UTODA n.a. n.a. Kigali Yes Rwanda Utilities Regulatory Agency 2004 n.a. Kinshasa Publicly owned buses only Operators/government 2006 No Lagos Yes (informally) Bus association 2005 n.a. Nairobi Yes Private bus association n.a. n.a. Ouagadougou Yes Direction Générale des Transports 2006 Yes Terrestres et Maritimes Source: Kumar and Barrett 2008. Note: n.a. = not applicable (because no large-bus operator exists); UTODA = Uganda Taxi Operators and Drivers Association. Urban Transport: Struggling with Growth 253 Fare Levels The average fare in the 14 cities is $0.31 for large buses and $0.25 for minibuses (table 6.12). The difference between the fares charged by large buses and minibuses is not consistent: in some cities (Addis Ababa, Bamako, Dakar), large buses appear more expensive than minibuses; in others, minibuses appear more expensive (Abidjan, Conakry); in Kigali, fares for the two services are equal. Unfortunately, the wide variation in fare structures, the difference between route structures of large and small buses, and the tendency of private companies to fragment their routes makes comparisons of real bus fares among cities very difficult. What can be safely deduced, however, is that because the operators are private and unsubsidized, the real fare for minibus services does cover costs, at least in the short term. In 12 of the cities (Addis Ababa and Nairobi are the exceptions), pas- sengers confirmed in the surveys reported by Kumar and Barrett (2008) that operators arbitrarily changed fares depending on circumstances such as bad weather and congestion. In Lagos, for example, an operator may increase the nominal fare of 40 to 70 (or even 90 on occasions). In Nairobi, on the other hand, drivers display fares on the inside of their windscreen, making opportunistic fair increases reportedly much less Table 6.12 Average Bus Fare (US$ per trip) City Large bus Minibus Abidjan 0.40 0.40­0.70 Accra -- -- Addis Ababa 0.25 0.12 Bamako 0.25­0.30 0.20­0.25 Conakry 0.18 0.21 Dakar 0.30 0.18 Dar es Salaam n.a. 0.16­0.24 Douala 0.30 n.a. Kampala n.a. 0.20­0.25 Kigali 0.28 0.28 Kinshasa 0.33 -- Lagos 0.40­0.56 0.38­0.39 Nairobi 0.25­0.40 -- Ouagadougou 0.30 n.a. Average 0.31 0.25 Sources: Various documents published by city authorities (data assembled by Kumar and Barrett 2008). Note: -- = not available; n.a. = not applicable. 254 Africa's Transport Infrastructure common. Being able to depend on consistent rates is important; when fares are variable, poor passengers often do not know whether they will be able to afford the bus fare home after work. Passengers also report other operating practices that increase uncertainty. For example, buses will sometimes shorten a trip to take advantage of a better commercial opportunity in the other direction. So-called short-turning strands passen- gers along the road, and operators rarely compensate the victims fully. The surveys suggest that fare uncertainty is a serious concern of many passengers. Affordability: Fares in Relation to Incomes Affordability of public transport obviously varies widely with passenger income and trip distance. For most people in Africa, and certainly for poorer people, transportation expenses are incurred primarily for the journey to work and a smaller number of equally essential purposes, such as trips for medical care. Transport is thus a necessity of life that must be provided for in the household budget, not a luxury. Anecdotal evidence from limited surveys shows that rising transport fares can isolate some people from employment opportunities, though this problem does not appear to be widespread. In most cases, budgets are tight in the region, and ridership drops sharply following fare increases, although it often rebounds after a few months. Most households reported some expenditure on urban transport, but a significant minority reported none. This is because in several of the cities under review, including Accra and Nairobi, low-income pockets can be found close to the city center (instead of on the outskirts, as in most devel- oping cities), and residents can meet their transport needs by walking. A standardized affordability index allows for comparisons across sample nations of the burden of public transport on the household budget. The index is based on the cost of 60 public transport trips per month--roughly equivalent to a daily journey to and from work. Data were collected from the most recent budget surveys available for each country and were analyzed at the city level. The total cost is expressed as a percentage of the monthly budget of an average household and of a first-quintile household in the 14 cities (see table 6.13). The results indi- cate that, averaged across all cities, the cost of 60 trips would absorb 8 percent of the monthly budget of an average household, but nearly one- third of the budget of the lowest quintile of households. Actual expenditures appear to be rather lower than the above analysis suggests, with transportation accounting for an average of 6.5 percent of Table 6.13 Spending on Urban Transport as a Share of Household Income Absolute monthly Percentage of expenditure on Percentage of average Percentage of Percentage of household budget transport for household budget first-quintile household households reporting spent on transport households with needed to pay for budget needed to pay positive expenditure for households with positive expenditure 60 one-way trips for 60 one-way trips on transport positive expenditure (US$) per month per month Abidjan 77 10.1 31.47 10.5 42.9 Accra 95 6.0 16.36 -- -- Addis Ababa 87 3.3 3.83 6.3 18.6 Bamako -- -- -- -- -- Conakry -- -- -- -- -- Dakar 92 4.3 15.08 3.1 11.3 Dar es Salaam 92 11.6 12.04 11.6 53.2 Douala 77 4.0 6.94 10.4 23.5 Kampala 81 7.4 13.08 7.8 41.0 Kigali 80 4.4 14.55 5.1 46.0 Kinshasa 49 2.8 5.43 10.1 31.0 Lagos 58 13.8 14.44 27.5 105.2 Nairobi 61 10.1 25.97 7.5 33.6 Ouagadougou 3 5.5 0.30 8.9 35.8 Average 90 6.5 13.29 8.0 32.7 Source: Kumar and Barrett 2008. Note: -- = not available. All data are for capital cities, except for Douala and Ouagadougou, where spending at the urban level nationwide is taken as a proxy for spending in the capital city. 255 256 Africa's Transport Infrastructure the budget of those households actually spending anything on transport in the 14 cities. There is very wide variation in this proportion: the share is just 3 percent in Addis Ababa but 14 percent in Lagos. When expressed in absolute terms, the amount that households spend on transport is much more consistent across cities, averaging $12­$16 per month. The exceptions are Addis Ababa and Kinshasa, where the expenditure is about one-third of that, and Abidjan and Nairobi, where it is about twice as much. In some cities (Abidjan, Dakar, Dar es Salaam, Kampala, and Kigali), there is a close correspondence between the actual budget share and the budget share needed to purchase the 60 trips. Elsewhere the difference is quite large and can run in either direction, with households spending sub- stantially more (Nairobi) or less (Addis Ababa, Douala, Kinshasa, and Lagos) than what is needed to purchase the 60 trips. In all cases, however, first-quintile households are at a disadvantage. They would typically need to spend 33 percent of their budget to pur- chase the 60 trips and in many cases a lot more, indicating that this level of mobility is completely unaffordable for the poorest households. Calculations of the same index for a number of Indian cities indicate broadly comparable results, with the average household needing to spend 5­10 percent of its monthly budget on the 60 trips, and with that share rising to 15­25 percent for the poorest households. The results presented here suggest that urban bus fares remain rela- tively high in relation to the purchasing power of the typical family, and very high in relation to incomes of the poorest. Nevertheless, even at low levels of expenditure, this translates into peak demand for around 200 seats per 1,000 residents, about five times higher than the supply available in any of the cities sampled. Financing and Subsidies Governments have tried to promote the use of larger buses, often with bilateral assistance from countries that manufacture them. The most suc- cessful experience seems to be in Addis Ababa, which has benefited from regular investment supported by bilateral assistance from the Netherlands and Belgium. In Kinshasa, the government-sponsored Société des Transports Urbains du Congo (STUC) has recently received a $33.5 million grant from the government of India to purchase new buses.8 In Lagos, the state government has even recently set up a bus company (Lagbus Asset Management Ltd.). Originally conceived as a Urban Transport: Struggling with Growth 257 way to gain the advantage of purchasing buses in bulk for local opera- tors using bilateral aid, the company has quickly converted itself into a government-subsidized operator. But these attempts have faced numerous problems, including noncom- petitive procurement of overpriced or inappropriate vehicles and insuffi- cient local technical support. In Accra, the government has established a commercial operator to revive large-bus services. The government has already procured several hundred new buses from China's Yaxing com- pany. The urban fleet in Douala received a similar upgrade in recent years, although in this case secondhand buses were imported from France. In both cases, a significant proportion is already out of service because of technical problems and difficulty obtaining spare parts. Meanwhile, arti- ficial suppression of fares on large public buses has led operators to defer the maintenance and replacement of aging fleets. Despite the good inten- tions of a number of governments, it still remains to be seen whether pub- lic funding of new buses for a publicly owned operator is a sustainable strategy in the African context. The private sector has done no better. Only in Nairobi does the private sector operate a significant number of large buses. The vehicles are com- posed of locally made bodies mounted on a truck-based passenger chas- sis. But a combination of government interventions in safety policy and competition from the matatus has made it more difficult to finance vehicle replacement; the average age of the fleet exceeded 12 years in 2005. Meanwhile, private financing of new larger vehicles is virtually nonexist- ent, partly because of the risk of losing capital investment to accident or theft--against which insurance costs would be prohibitive--and partly because of the longer payback periods for new and more expensive vehi- cles relative to secondhand, less expensive vehicles. In Dakar, the World Bank has supported an innovative scheme in which operators pool funds to commit to collectively finance bus investments. The scheme has yet to prove sustainable, though 505 buses were bought, as projections suggest that companies will need to charge higher fares to fully recoup their investments. In general, it would appear that the fragility of the regula- tory environment will continue to discourage the private sector from investing in large buses. Finding private financing for smaller vehicles has also proved difficult. The private sector has proved capable of raising funds through commer- cial borrowing only to purchase the most basic secondhand minibuses, for which payback periods are relatively short (box 6.2). Investors are often limited to using family savings for capital, often in the form of interest-free 258 Africa's Transport Infrastructure Box 6.2 Financing Large vs. Small Buses in Nairobi The average price of a secondhand 14-seat matatu (five to seven years old) is approximately $11,800. This may be borrowed commercially, but it is more often assembled from family sources. The net return to the owner is about $21 per day after deductions for insurance, license, vehicle inspection, the cost of tires, and regular maintenance, all of which total about $6,200 per year. Based on these fig- ures, the owner would recoup the capital cost in about two years, well within the average working life of the vehicle. By contrast, a new, locally bodied, 35-seat matatu built on a light truck chassis costs about $46,000. The purchaser normally would have to borrow from a com- mercial investor, who would insist on comprehensive insurance at a cost of about $9,200 a year, much more than the insurance normally used to cover smaller vehicles. The cost of insurance alone makes large new vehicles unprofitable. The Matatu Owners Association estimates that many who bought bigger matatus in 2006­08 are likely to go bankrupt. Source: Kumar and Barrett 2008. loans from family and friends. Bank finance is rarely available because the banks are reluctant to accept the vehicles as collateral in the absence of a secure secondhand market, and because revenue streams in the informal sector are too unreliable to assure the banks of repayment. Subsidies allow operators to charge fares that are low enough to be affordable for low-income passengers, while still providing sufficient capacity and quality of service to accommodate demand. In this respect, transport operations are a commercial enterprise, and the state reim- burses the operator for providing discounts to certain customers (usually students or the aged). In many cases, however, governments have kept fares low but failed to pay the promised subsidies. This practice inevitably leads to deterioration in services--first in quality (because of reduced maintenance) and then quantity (because of an inability to finance fleet renewal) (Gwilliam 2000). Large buses benefit from some form of public support in most of the 14 cities that have large-bus services. The only exception is Nairobi, where the operator KBS has full control over setting fares. In Accra, large-bus fares are too low for operators to fully recover their operating Urban Transport: Struggling with Growth 259 costs over the life of a bus. The government-sponsored operator Metro Mass Transit Ltd. (MMT) is facing severe cash-flow constraints and will soon require subsidies to remain solvent.9 The government of Addis Ababa subsidizes bus fares in the city even though the operator Anbessa is federally owned. The government of Senegal has accepted that in Dakar, the private large-bus company Dakar Dem Dikk (DDD) will require a subsidy to operate at the regulated fare. Douala has a history of subsidizing fares on large buses or restructuring operators. In 2008, the Douala Urban Council bought 38 percent of the shares of the private company, Société Camerounaise de Transports Urbains (SOCATUR), which had previously taken over from the defunct state company, Société de Transports Urbains du Cameroun. The SOCATUR now operates under a contract with the government signed in December 2008. The Société de Transport Abidjanais (SOTRA), the large-bus operator in Abidjan, receives an annual subsidy that allows it to offer reduced fares to civil servants, the military, and students. Finally, in Ouagadougou, the government exempts the SOTRACO from tax and duty on fuel, tires, and other imports. Beyond direct operating subsidies, operating companies also benefit from soft loans on better-than-normal commercial terms to finance the acquisition of buses (Accra, Addis Ababa, Dakar, Douala, and Lagos). Deficit financing of bus operators can also lead to inefficiency in the organization that receives the subsidy. Even with buses averaging five years in age, the large-bus operator in Addis Ababa can achieve only 83 percent fleet availability--well below what an efficient operator should be able to achieve (90­95 percent). The short-lived, publicly owned operator in Nairobi was unable to compete commercially despite its access to duty-free inputs and donated vehicles. The operator in Accra reported a 40 percent "leakage" of revenues, indicating that the main ben- eficiaries of subsidies in this case were the operator's employees. Subsidies through government capital grants for new vehicles can also undermine efficiency. Accra's buses have not been subjected to any formal specifica- tion or a transparent procurement process, without which they are unlikely to offer the lowest possible cost of operation over their life cycle. The latest public procurement in Lagos is reported to have increased the cost of the chosen vehicle by 60 percent over the price that private oper- ators would expect to pay. In Addis Ababa, bilateral concessional aid has financed the importation of buses built in Europe, such that the main beneficiary is actually the European builder and not the local bus system (a situation not unlike colonial times). 260 Africa's Transport Infrastructure In sum, the experience of financing bus operations differs widely, but there still appears to be no viable standard model for Africa comparable to the competitive tendering of franchises, which has worked well in many European countries. Regulation The regulatory framework for urban public transport typically comprises several elements, including service planning, controlling entry into the market by new operators, allocating routes to market participants, licens- ing vehicles and drivers, establishing procedures for vehicle inspection, and setting passenger fares and tariff structures. These functions are often dis- persed among a number of agencies, both local and national (table 6.14). The Gulf between Principles and Practice In all 14 cities, commercial vehicles must be registered and licensed to carry passengers. Vehicles are inspected for roadworthiness at the time of registration and annually or semiannually thereafter. Inspection standards are often outdated, however, and fail to address environmental concerns, such as emissions and noise. Drivers of commercial passenger vehicles must also pass a test and obtain a special license. In most cities, the driv- ers of large vehicles must be more experienced and more highly qualified, while minibus drivers require no qualifications beyond those needed to operate a private car. Like vehicle inspections, driver training and testing are relatively weak, and most driving schools and testing stations do not have a full-size bus. Some of the 14 cities are taking measures to limit drivers' ability to bypass testing procedures by obtaining fraudulent doc- uments or altering legitimately issued ones. For example, Addis Ababa is now introducing a more secure, counterfeit-proof system. For the formal bus sector, many countries retain institutions and processes inherited from a colonial past, whether British or French. For example, in Francophone West Africa, the large-bus services in Abidjan, Dakar, and Douala are, in principle, tightly regulated, with a ministry of transport allocating routes to a monopoly supplier that charges specified, controlled fares. Large-bus services have well-defined route structures that are sometimes out of date but could in principle be revised to reflect population growth and movements. Some of the old formal arrangements remain. In Douala, a transport- monitoring commission (Comité d'Organisation et de Suivi de Transports Urbains) operates in each municipality, with members representing Table 6.14 Regulatory Framework for Urban Public Transport Legal restrictions System for allocation of routes on entry of buses Entity responsible for Entity responsible for vehicle City to operators to the market licensing of vehicles inspection Abidjan Existing concessions for large-bus Yes for large AGETU SICTA (private concessions) operations buses Accra None None Local authority None Addis Ababa Transport authority None Department of Trade and Industry Federal Transit Authority Bamako District and municipality None Direction Nationale des Transports Yes Terrestres et Fluviaux Conakry None None Ministry of Transportation None Dakar Ministry of Transportation None Ministry of Transportation CETUD Dar es Salaam SUMATRA None SUMATRA Ministry of Public Safety and Security (traffic police) Douala Comité d'Organisation et de Yes n.a. None Suivi des Transports Urbains Kampala None None Licensing board None Kigali None None Rwanda Utilities Regulatory ONATRACOM Agency (regulator) Kinshasa None None n.a. None Lagos None None Motor Vehicle Administration LAMATA Nairobi Transport Licensing Board None Transport Licensing Board None Ouagadougou None None Direction Générale des Transports None Terrestres et Maritimes Source: Kumar and Barrett 2008. Note: n.a. = not applicable; LAMATA = Lagos Metropolitan Area Transport Authority; ONATRACOM = Office National de Transport en Commun; SICTA = Société Ivoirienne de Contrôle Techniques Automobiles et Industriels. 261 262 Africa's Transport Infrastructure each of the relevant ministries and the operators. But decisions on the route structure, licensing, and fares are made by the ministries of transport and finance in Yaoundé. In Dakar, in principle, the Ministry of Transport controls route licensing, partly to protect the new bus company, DDD, from competition from cars rapides. In practice, however, government-granted route monopolies have not protected DDD. Effective allocation of routes to operators appears, in effect, to be controlled by the operators' syndicates. In Abidjan, the SOTRA operates under the technical supervision of the Ministry of Transport and the financial supervision of the Ministry of Economy and Finance, with a monopoly right to operate public transport service in a defined territory within the city. But, as noted, most of the public companies covered by the traditional regulatory regimes have languished or failed, supplanted by minibus services operating outside the formal regulatory system. In most cases, rational public planning and administration of the route structure is non- existent, the replacement of formally operated large-bus services with informally operated minibuses having eliminated any vestige of strategic public control. Even where large-bus services still exist, they are a small part of the total supply of public transport, and not components in a well-planned integrated network. Permits are routinely issued on request, without consideration of sup- ply and demand in the city. Governments may also allocate vehicles to routes, as in the SUMATRA's allocation of the dala-dala routes in Dar es Salaam. But the allocation of vehicles to routes is rarely enforced, except in Nairobi. So, in practice, operating permits are valid throughout the jurisdiction of the issuing authority and are recognized by adjoining juris- dictions within metropolitan areas. This may not be a bad thing. Using permits to assign routes makes sense only if the issuing authority has a good understanding of the transport network and the changes needed to better accommodate passenger demand. In general, however, the 14 cities do not have sufficient understanding of the network to properly manage it. In Addis Ababa, while the transport authority at present issues route licenses, it intends to devolve licensing responsibility to the operators' associations. The lack of formal regulation is offset by the existence of unions, associations, or syndicates that organize the activities of the sector and provide a degree of self-regulation. Almost all of the 14 cities have at least one syndicate that performs this function. In Kampala, membership Urban Transport: Struggling with Growth 263 in the Uganda Taxi Operators and Drivers Association (UTODA), the main industry union, is obligatory and may be violently enforced. The general practice is for the syndicates to collect dues from their members, who then have the right to use the terminal facilities man- aged by the syndicate. The syndicates also charge daily fees based on terminal use. A charge is normally paid on first use of the terminal each day, and this may then be supplemented by individual departure charges, sometimes based on the number of passengers carried, and also by further charges at the destination terminal and at major stops along the route. Some syndicates play a role in regulating routes and setting fares. Self-regulation has created an orderly market and mitigated the worst consequences of unbridled competition on the road. Route terminals are well managed, and overloading and fare gouging are uncommon. Members who flout the rules are subject to disciplinary action. Nevertheless, union control has drawbacks: since the routes run between union-controlled terminals, the route network is overly rigid and gener- ally fails to match transport supply to passenger demand. As a result, too many passenger trips involve one or more bus changes, which increase trip cost and length. Inadequate Enforcement All 14 cities lack sufficient institutional capacity and integrity to prop- erly enforce vehicle standards. In Lagos, 37 percent of vehicles were operating without a valid certificate of roadworthiness, and 47 percent without a valid test certificate. In Accra and Addis Ababa, between 30 and 50 percent of vehicles were operated by unlicensed drivers. In both Abidjan (SOTRA) and Kinshasa (STUC), however, buses were required to undergo regular inspections by professional mechanics in dedicated garages, and bus drivers and conductors were professionally trained. Tata buses owned by the STUC benefited from technical assis- tance from Tata Motors Ltd. Until recently, Accra had only four vehicle inspectors, whose duties also included accident investigation. The city now has 14 inspectors, which is still not enough. The prevalence of unsafe vehicles is well known. In each of the cities surveyed, passengers had nick- names for unsafe vehicles, such as "DMC" for "dangerous mechanical condition." To address the failure of the existing inspection regime, Lagos has introduced new tests for vehicles that are more than five years old. 264 Africa's Transport Infrastructure The tests are carried out at licensed private testing stations, but so far, stan- dards of integrity remain unsatisfactory. Both Addis Ababa and Kampala intend to privatize their vehicle inspection regimes. In Accra, all vehicles intended for commercial passenger transport must be registered when they are first imported, converted for passen- ger carriage, or resold. At the time of registration, the vehicle must be tested for roadworthiness by the Driver and Vehicle Licensing Authority and then retested every six months thereafter, which is twice the frequency required for private vehicles. The institutional and techni- cal capacity of the test centers, however, limits the effectiveness of the testing regime. The Driver and Vehicle Licensing Authority is also responsible for driver testing, the quality of which is dubious. Licensing requirements for owners of commercial passenger vehicles are either weak or nonexistent, which give transport operators little incentive to raise their standards. A stronger regime would require operators to maintain their vehicles in roadworthy condition and would enforce the requirement through inspections, tests, and sanctions. None of the 14 cities has such a system, and enforcement problems are common. In a survey in Lagos, 21 percent of drivers interviewed acknowledged that they did not hold a valid driver's license. Drivers commit many trans- gressions throughout the day, for which they pay petty bribes to police officers. Drivers can work long hours--to a point that exceeds the safety threshold. In Kenya and Uganda, for example, driver shifts average more than 12 hours a day for 6 or 7 days a week, although driving hours are normally closer to 7 or 8 hours. But police interviewed for this study were generally unconcerned about drivers' hours of operation, since off- peak periods allow drivers to rest. Conditions appear to be less stressful in West Africa, where cars rapides normally have two drivers (who both work 8-hour shifts), a conductor, and a route assistant. Safety The highest road fatality rates (deaths per 10,000 motor vehicles) worldwide occur in African countries, in particular Ethiopia, Uganda, and Malawi, as shown in chapter 2 and appendix 2j. It is estimated that between 20 and 40 percent of fatalities, and a larger proportion of nonfatal injuries, occur in urban areas.10 A high proportion of fatali- ties and injuries involve vulnerable pedestrians. For example, in urban areas of Zambia, pedestrians account for two-thirds of fatalities and over half of all road traffic injuries, compared with 30 percent of fatalities Urban Transport: Struggling with Growth 265 and 12 percent of injuries in rural areas. Similarly, in Ethiopia, pedes- trians represented 85 percent of all injuries within Addis Ababa but only 40 percent nationwide (Downing and others 2000). Nevertheless, the interests of pedestrians continue to be neglected both in road infrastructure design and in traffic management and enforcement. Proper institutional recognition of the problem is a good starting point. The establishment of a management unit at a high level of government has proved effective at the national level in Ghana, and at the municipal level in other parts of the world (for instance, in the Brazilian capital). Comprehensive programs have already had some success in KwaZulu- Natal, South Africa. Because of its dominance in urban traffic, and in view of the condi- tions in which it operates, public transport plays a large part in urban traffic accidents. This has been specifically recognized in Kenya, where a 2003 law requires all public service vehicles to drive no more than 80 km per hour and to provide seat belts to all passengers. This law reduced passenger loadings and vehicle speed for both conventional large buses and matatus, which in turn reduced the financial viability of the services and put pressure on fares. The safety measures were reported to have reduced road fatalities by 40 percent and serious injuries by 50 percent in the first year of their operation (Chitere and Kibua 2005). This experience suggests that, perhaps with more careful design, safety measures can reduce road traffic injuries and fatalities in African cities. The Way Forward The picture of urban transport that emerges from the studies is one of inadequate and poorly managed infrastructure used by an insuffi- ciently regulated vehicle fleet. As a result, formally operated modes of urban transport have given way to informally operated minibuses, which have worsened congestion and raised concerns about fare afford- ability for the poor, safety, and environmental standards. Regulation of public transport service has either failed or, worse, has bankrupted the large-vehicle operators. While minibus operators' associations often provide some degree of self-regulation, their primary goal is to maxi- mize revenues, not to serve the interests of their customers. Improving urban public transport in Africa is crucial, with several priorities to be addressed. 266 Africa's Transport Infrastructure Priority 1. Developing an appropriate metropolitan government structure Many of the difficulties that the urban transport sector has faced can be attributed to a lack of strategic planning by a central authority. This can be addressed by a number of steps, including the following: · Identifying and defining metropolitan areas within which spatial interactions (such as major commuting movements) are significant (this could relate to all public services or more narrowly to transport issues) · Reassigning responsibility for all strategic transport issues (such as road and public transport network planning, traffic management, and public transport fare and service policy) to the metropolitan authorities · Establishing financing arrangements for metropolitan services to ensure that all those who benefit from metropolitan-level services contribute to the costs of provision. Priority 2. Establishing a sustainable financial basis for urban roads The inadequate quantity and poor quality of urban roads is commonly attributed to the inadequacy of funding for roads at the urban level. Two steps can be taken to address this funding problem: · Reserve an adequate share of road-fund revenues for urban roads. This can be done either by formula or by including urban road interests in the road board decision process. · Identify new sources of funds for urban roads. This could be achieved either by earmarking existing road-based taxes, such as vehicle- and driver-licensing revenues and parking charges, or by developing new taxes, such as congestion charges. Priority 3. Reducing road congestion Congestion has commonly been attributed to a combination of financial and institutional weaknesses. Cities can take the following actions to address these weaknesses: · Strengthen the technical competence of the metropolitan-level authority to enable it to manage effectively in a range of relevant areas such as road finance and traffic management. · Establish strong traffic management agencies to ensure the adoption of strong traffic restraint rules and procedures. Urban Transport: Struggling with Growth 267 Priority 4. Reestablishing a public transport regulatory framework The proliferation of the minibus has occurred by accident rather than design, largely as a result of unrealistic fare and service obligations placed on the public sector and large-bus operators, unaccompanied by financial support. As large-bus companies fail, informal operators become respon- sible for the majority of urban public transport in the region. Cities must therefore do the following: · Redefine or clarify public service vehicles to include all vehicles carry- ing passenger at separate fares, irrespective of size or type, so that the informal sector can be aligned with public service regulation. · Establish public oversight of regulation enforced by operators' associ- ations. · Enforce the loss of operators' rights for public service providers that fail to meet the obligations included in their licenses or contracts. Priority 5. Developing a comprehensive strategy to reduce road accidents Despite the paucity of reliable statistics, there is little doubt that road accident rates in African cities are unacceptably high. Given what is known about the characteristics of the victims (largely pedestrians), the location of accidents (on links rather than at junctions), and the vehicles involved (often public transport vehicles), a range of potentially benefi- cial measures is known but frequently not implemented. A comprehen- sive strategy would thus need to include the following: · Creation of a road safety unit at a high level of municipal or city government · Better allocation of space for pedestrians in urban road design · Better separation of commercial and traffic activities, if necessary by provision of attractive locations for hawkers off main roads · Better enforcement of safety requirements for public transport and freight operators, including controls on speeding and overloading. Notes 1. The main source document for this chapter is Kumar and Barrett 2008. 2. The data were collected between June 2004 and December 2006 for studies by the Public Private Infrastructure Advisory Facility (PPIAF), the Sub- Saharan Africa Transport Policy Program (Gleave and others 2005), and the World Bank. 268 Africa's Transport Infrastructure 3. An electric multiple unit, or EMU, is a multiple-unit train consisting of more than one passenger carriage; all carriages in the train carry passengers, using electric traction motors incorporated within one or several of the carriages. 4. The modal share describes the percentage of trips undertaken using a partic- ular type of transportation. 5. The colloquial names sometimes refer to a fare. For example, in Ghana "tro- tro" means "three pence-three pence," and refers to the fare set in the early 1970s to undercut the government-set fares. Similarly, in Kenya the word "matatu" is derived from the local term "mang otore matatu," meaning "thirty cents," the standard fare once charged. In other cases, names refer to the vehi- cle type. In Nigeria "molue" means "molded," the vehicle being a molded body on a truck chassis. "Danfo" means "stands on its own" and refers to an integral- construction microbus. 6. Such vehicles are built with the body and underframe as an integral unit (instead of with the chassis and bodywork as separate constructions). 7. The SOTRACO is a limited liability company, created in 2003 to succeed the defunct Société de Transports Alpha Oméga, which wound up its oper- ations in 2000 after encountering serious financial difficulties. Its objectives were to develop a public-private partnership involving experienced private operators in the transport sector in Ouagadougou, and to delegate manage- ment responsibilities to private shareholders. The commune of Ouagadougou owns 15 percent of the SOTRACO, and the private operator 85 percent. The government subsidizes the SOTRACO by exempting it from import duties and taxes on fuel and tires. 8. STUC is a French-style société d'économie mixte, created in 2004 with a man- date to find international partners to provide urban transport in Kinshasa and other cities of the Democratic Republic of Congo without any charge on the national budget. It started operation in February 2006 with 10 buses. In June 2006, it received 228 new Tata buses financed by the Indian government. 9. MMT was incorporated in 2003. The shareholders include State Insurance Company, National Investment Bank, Ghana Oil Company Ltd., Agriculture Development Bank, Prudential Bank, and the Social Security and National Insurance Trust. These together have 55 percent shareholding. The govern- ment of Ghana holds the remaining 45 percent of shares. 10. Based on discrepancies between hospital statistics and official police accident statistics, it is believed that there is significant underrecording of all accidents, in particular of nonfatal accidents and accidents involving women. In 2009, the United Nations Economic Commission for Africa (UNECA) undertook an extensive review of road accidents in Africa, which is due to be published late in 2010. Urban Transport: Struggling with Growth 269 References Bullock, R. 2009. "Railways in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 17, World Bank, Washington, DC. Chitere, P. O., and T. N. Kibua. 2005. "Efforts to Improve Road Safety in Kenya: Achievements and Limitations of Reforms in the Matatu Industry." Institute of Policy Analysis and Research, Nairobi, Kenya. Downing, A., G. Jacobs, A. Aeron-Thomas, J. Sharples, D. Silcock, C. van Lottum, R. Walker, and A. Ross. 2000. Review of Road Safety in Urban Areas. Crowthorne, Berkshire, U.K.: TRL. Gleave, G., A. Marsden, T. Powell, S. Coetze, G. Fletcher, I. Barrett, and D. Storer. 2005. "A Study of Institutional, Financial and Regulatory Frameworks of Urban Transport." Sub-Saharan Africa Transport Policy Program Working Paper 82, World Bank, Washington, DC. Gwilliam, K. M. 2000. "Public Transport in the Developing World." Transport Series Discussion Paper, World Bank, Washington, DC. Gwilliam, K., V. Foster, R. Archondo-Callao, C. Briceño-Garmendia, A. Nogales, and K. Sethi. 2009. "The Burden of Maintenance: Roads in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 14, World Bank, Washington, DC. Kumar, A. J., and F. Barrett. 2008. "Stuck in Traffic: Urban Transport in Africa." Africa Infrastructure Country Diagnostic Background Paper 1, World Bank, Washington, DC. Mobereola, D. 2008. "Africa's Megacity Needs an Urban Rail Backbone." Railway Gazette International, November 14. UITP (International Association of Public Transport). 2001. Millennium Cities Database for Sustainable Transport. UITP, Brussels. World Bank. 2002. Cities on the Move: World Bank Urban Transport Strategy Review. Washington, DC: World Bank. CHAPTER 7 Spending to Improve Connectivity The foregoing chapters on the major transport modes and on urban trans- port identified deficiencies in infrastructure capacity, quality, and condi- tion that call for investment if they are to be remedied. They also identified circumstances in which maintenance spending must be increased, in particular for roads. But they did not convert those require- ments into specific spending requirements. That is the purpose of this chapter.1 In theory, investment requirements can be computed by summing the costs of all feasible investment projects--that is, the capital costs of all investments that show a positive net present value at a discount rate equal to the current cost of capital. But to make such an estimate would be an enormous, perhaps impossible, task. Even if it could be done, the result would, in its putative precision, be misleading. For one thing, the computation of the costs and benefits of regional projects would be viti- ated by the political realities of present-day Africa. In place of that approach, therefore, this chapter offers a model designed to estimate the cost of achieving specific connectivity targets considered appropriate to the African context without going through the exercise of appraising every project that might be found feasible based on a full economic analysis. 271 272 Africa's Transport Infrastructure The Expenditure Model in Brief The model presented here is based on the concept of connectivity, which presumes that the purpose of transport infrastructure is to facilitate transport services that connect people with one another, with markets, and with the social services that are available in urban centers. The extent and quality of transport infrastructure networks and facilities, and the standard to which they are maintained, are obviously critical to the notion of connectivity, as poor or inadequate infrastructure cannot provide a reliable base for reasonable access to people, places, markets, and services. The connectivity model estimates the total cost of providing specified amounts of infrastructure at a specified standard to meet a specified degree of connectivity over a certain period of time. The cost estimates generated by the model are very sensitive to these specifications and thus, ultimately, to the targeted degree of connectivity that is to be achieved. In the model, each combination of specifications is called a scenario. The model, as it is described here, provides a reference set of parameters that make up what is called a base scenario. The estimated costs of this base scenario are the standard against which the costs of any other scenario may be compared. The base scenario adopted here incorporates a level of connectivity in principle comparable to that of developed countries, with all facilities maintained in good condition.2 Such a scenario would be out of reach for many countries--which is no doubt why the ambitious targets articulated in many national plans so often go unrealized. For that reason, the model permits the specification of less ambitious alternatives. The pragmatic sce- nario described in this chapter is one of many potential lower-cost alter- natives to the base scenario. It aims for a somewhat lower level of connectivity, with infrastructure maintained in only fair condition. The scenarios are described in more detail in the section on the application of the model. The model is presented in three stages. The first stage comprises inputs--that is, the scenario specifications and data needed to run the model. In the second stage, the costs of the base and alternative scenar- ios are calculated. The third stage comprises outputs, in the form of comparisons of the costs of the base and alternative scenarios. The rela- tionships between the different components of the model are illustrated in figure 7.1. Figure 7.1 A Model for Estimating the Costs of Transport Infrastructure Expenditure Needs Inputs Model Outputs desired changes in: databases investment needs connectivity regional national by country rural base alternative individual countries urban scenario country groups scenario definition definition Africa total condition good regular standards macroeconomic by expenditure category roads : surface type and width submodel improvement airports : runway length upgrade ports : types of berths expansion railways : axle load maintenance sources of macroeconomic database infrastructure for connectivity standards by connectivity and mode current infrastructure regional database road airport port railway unit cost total scenario cost calculations national matrix road airport rural urban scenario comparisons and integrations 273 Source: Carruthers, Krishnamani, and Murray 2009. Note: SSA = Sub-Saharan Africa. 274 Africa's Transport Infrastructure The first stage begins with a specification of the three input compo- nents of each scenario for which costs are to be estimated: · Connectivity targets for regions, countries, rural areas, and urban areas · The initial capacity and construction of the infrastructure to be pro- vided, referred to here as the infrastructure "category" (for example, a two-lane paved road or a one-lane gravel road) · The condition in which the infrastructure will be maintained. The inputs also include a macroeconomic database that provides eco- nomic and social projections needed to calculate scenario parameters. For example, information on international trade volumes is needed to estimate the number of port berths required for handling that trade. Other data- bases quantify existing transport infrastructure and the costs of improving, upgrading, and expanding that infrastructure to reach the specified levels of connectivity. In the second stage, the model draws on various databases to calculate the links needed to achieve the connectivity targets articulated in the first stage. It then reviews the database of existing infrastructure to see how much of what is needed can be found there, irrespective of its current cat- egory or condition. The options are as follows: · To use the existing infrastructure in its current category and condition · To improve its condition (for example, from poor to good) · To upgrade its category (for example, from a gravel road to a paved road) · To build new infrastructure. Each item of infrastructure, such as a link in a road network or a run- way at an airport, is considered in light of these options. The model com- putes the costs of each option by applying a matrix of unit costs to each type of activity (for example, improving a road from poor to good qual- ity and upgrading its category from gravel to paved). Although the model calculates the costs of these activities sequentially, in practice they are usually implemented simultaneously. This second stage of the model yields the cost of meeting the standards specified in the base and alternative scenarios. That cost can be expressed as a total cost over a 10-year period, as an average annual cost, or as a per- centage of the gross domestic product (GDP) per year. Spending to Improve Connectivity 275 The third stage of the model offers a disaggregated presentation of the costs of the base and alternative scenarios in three ways: · By country, by groups of countries, and for the region as a whole · By type of expenditure--differentiating among spending to upgrade and improve existing infrastructure, to provide new infrastructure, and to maintain all infrastructure · By type of connectivity (regional, national, rural, and urban) and trans- port mode (road, rail, airport, and port) The costs can be calculated for any combination of the three cate- gories. For example, the model can estimate the cost of maintaining the roads used for regional connectivity in low-income, fragile countries. A Detailed Look at the Model's Inputs As noted in the previous section, the inputs to the model are (i) data from five databases described below and (ii) specified requirements, or targets, for connectivity, infrastructure standards, and infrastructure condition. Five Databases Five main databases were used in the study. For more detail on these data- bases, see Carruthers, Krishnamani, and Murray (2009). Macroeconomic database. To estimate what has to be done to meet a connectivity target, the model requires projections of certain macroeco- nomic and social variables. Regional connectivity and national connectiv- ity, for example, are defined in terms of air and road links between cities of various sizes. The macroeconomic database included the following figures for each country: · Current (as of 2008) and projected national population · Current and projected urban population · Current and projected total GDP · National productive land area · Urban land area · Current and projected international trade, disaggregated by imports and exports and by containerized, general, and bulk freight. 276 Africa's Transport Infrastructure The projections of city populations were taken from a widely used and readily available database of country and city populations (Demographia 2009) that is, in turn, based on United Nations estimates of the growth of urban populations. Because GDP projections are generally not available by country for periods of longer than three or four years, estimates were based on short-term projections for African countries done specifically for this analysis. Geographic information system. The connectivity approach was under- pinned by an extensive geographic information system (GIS) data plat- form assembled specifically for the Africa Infrastructure Country Diagnostic. The data platform included as much of the current transport infrastructure as possible, including interurban road and rail networks, airports, and ports. Used primarily to calculate the distance (in transport network kilometers [km]) between the geographic and demographic features of interest in each country, it includes geographical databases covering the spatial distribution of population, administrative bound- aries, geographic and environmental features, and GIS references for all towns and cities of more than 25,000 people. To determine an appro- priate standard for urban connectivity, the GIS platform was comple- mented by detailed databases on city population and population density, discussed above. GIS-referenced locations for cities and towns were drawn from the Global Rural-Urban Mapping Project (GRUMP) of the Center for International Earth Science Information Network (CIESIN) at Columbia University's Earth Institute. The survey of road and rail infrastructure was based on references from the Digital Chart of the World. Road surface and width were determined manually from Michelin's regional map series of 2004 (1:4,000,000 scale). Locations for ports were based on the GIS data set available from the U.S. National Geospatial-Intelligence Agency (NGA), which publishes a World Port Index containing information on world ports and related facilities. GIS references for airports were shown on the same large-scale Michelin maps consulted for roads.3 Roads. The road network was assessed using a data set compiled for a previous World Bank study of road investment needs in Africa. That data set drew on spatially integrated data compiled by CIESIN. Other data for the roads database were drawn from Gwilliam and others (2009) and from national statistics on classified and unclassified road networks, as well as from various reports of the Sub-Saharan Africa Transport Policy Program (SSATP). The latter were particularly useful for data on urban roads. Various SSATP sources were also used for data on the condition of Spending to Improve Connectivity 277 interurban and urban roads, and recourse was also made to many publi- cations by national road agencies. Other infrastructure data. Several additional sources were used for transport infrastructure data. Bofinger (2009) was the main source for information on airports and runways. Mundy and Penfold (2009) pro- vided ports data, supplemented by data from Web sites maintained by national port agencies. The World Bank's railways database (World Bank 2010) was a main source for the length and gauge of national railway net- works. Data on the length of operating routes that had long-term prospects for sustainable investment were based on Bullock (2009). Unit costs. The database of unit costs covered all the construction and maintenance operations that might be needed to meet the specified connectivity targets. The source of the unit costs for road-related con- struction and maintenance was the same as that used in Gwilliam and others (2009). The unit costs for railway activities were based on those used in a selection of recent World Bank railway projects, checked against the cost data provided in Bullock (2009). Port unit costs and air- port investment costs were provided by port and airport specialists at the World Bank. The cost estimates for the connectivity targets are very sensitive to unit costs. As applied in this study, the model uses the same unit costs for all countries. In practice, however, the costs are known to vary substantially. For example, while the median estimated cost of rehabilitating a two-lane road in Africa in 2006 was $300,000 per kilometer, the upper quartile esti- mate was over $450,000. Moreover, the costs vary across countries, and are likely to be considerably above the median in landlocked countries. The implication is that the costs may be underestimated for some of the poor- est countries--so that their needs are even greater than estimated in this chapter, and their shortfall greater than that estimated in the next chapter. This limitation can be overcome. The database design allows the costs to be scaled up or down for a specific country or group of countries, for all unit costs or just for those related to a particular transport mode or a con- struction or maintenance operation. For more detailed application to any individual country, it would be necessary to calibrate the cost model on local data. The cost matrix is structured by type of operation (improving the con- dition of infrastructure, upgrading its category, building new infrastruc- ture, or performing maintenance) and by mode (regional roads, regional ports, regional railways, regional airports, national roads, national airports, rural roads, and urban roads). 278 Africa's Transport Infrastructure Standards of Regional, National, Rural, and Urban Connectivity The connectivity approach consists of identifying the key geographic and demographic features of each country and then quantifying the transport infrastructure needed to connect those features. Features of international interest (such as capital cities, deepwater ports, and international borders) provide connectivity across the entire region, whereas features of purely national interest (mainly secondary cities and provincial capitals) provide national connectivity. In rural areas, the focus is on connecting agricultural land to markets, while in urban areas the focus is on connecting house- holds to the activities that are necessary for their well-being. The model therefore considers four types of connectivity: regional, national, rural, and urban. Regional connectivity. Regional connectivity is needed for the interna- tional movement of people and goods. The assumption is that connectiv- ity requires a road network that links national capitals to one another, to all other cities with a population of more than 250,000,4 to international land borders, and to deepwater ports. In the model, interurban road infra- structure is specified in terms of the actual road carriageway, without regard for bridges, tunnels, and other structures that would require a level of geographic specificity and knowledge that is beyond the scope and resources of the study. Because of this limitation, the estimates presented here are likely to underestimate the full cost of the reference standard of connectivity. Another assumption is that larger cities and national capitals require airports with appropriate runways and terminals, and that current termi- nal capacity is sufficient to deal with current demand. Cost estimates therefore include only the additional area needed to satisfy projected increases in demand. To ensure regional connectivity, each country must have access to a deep-sea port of regional significance (not necessarily within the country) with appropriate container, general freight, and bulk freight berths of depths suitable for the current generation of ships. Out of the current rail network of more than 62,000 km, only about 55,000 km are currently in operation. Low-volume lines--those with less than 1 million net tonnes of freight per year--are very unlikely to merit full rehabilitation, and lines carrying bulk traffic will normally gen- erate greater (and more certain) benefits than those carrying general traf- fic. The result is that only about 29,500 km of line have a chance of justifying the investment needed to keep them in operation--and only in unusual circumstances could keeping them operational contribute to Spending to Improve Connectivity 279 connectivity standards or provide cost-effective alternatives to road transport of freight. Nevertheless, because future demand for rail infra- structure is very difficult to assess, the rail connectivity standard keeps those 29,500 km of railway lines in operation and upgrades them to an 18 tonne or 25 tonne axle load depending on the volume of freight they handle. National connectivity. National connectivity is provided by a transport network that links each country's provincial capitals to other medium- sized cities, defined as cities having a population of at least 25,000. The network is used for nationally traded goods, for access to services such as health and education, and for access to family members in other cities. A national connectivity network is generally limited to roads within a single country, except in a few cases where a lower-cost route between a medium-size city and the existing national network might overlap with a cross-border regional road or with the national network in a neighboring country. Medium-size cities with populations of more than 100,000 were assumed to require airports with shorter runways (1,524 meters) and terminals with slightly lower capacity standards (16 square meters [m2] per domestic passenger) than those needed for regional capitals. No additional railway or port facilities were assumed beyond those required for regional connectivity, the national require- ments being fully satisfied by the regional infrastructure (provided that the infrastructure was up to standard). Rural connectivity. Two approaches to estimating rural accessibility or connectivity were explored. The first is centered on a social criterion for connecting people, the second on a commercial criterion for connecting rural production to national and international markets. After considera- tion, the second approach was chosen. The social measure of rural accessibility is based on the Rural Accessibility Index (RAI) discussed in chapter 2. Roberts, Shyam, and Rastogi (2006) have developed an index that measures the share of the rural population living within 2 km of a road that is passable in all weather. In the world's middle-income countries, 94 percent of rural people meet that criterion. But in African countries, the share is only 34 percent, with values ranging from 5 percent in Sudan to 67 percent in Lesotho. The estimates of the current RAI values are based on household survey results and extrapola- tions from respondents' perceptions of whether they live within 2 km of a paved road. It is impossible to know which roads they take into account 280 Africa's Transport Infrastructure when reporting their assessments to survey takers. But from the GIS data available on population distribution and existing roads, it was possible to estimate the road system necessary to satisfy different RAI targets. A tar- get of 75 percent of the rural population living within 2 km of a single- lane road with a single surface treatment was taken as the base scenario, and a target of 50 percent living within 2 km of an improved gravel road with drainage was adopted for the pragmatic scenario. The agricultural output measure of rural connectivity is the alternative approach; it, too, was introduced in chapter 2. The criterion used for both scenarios was that of connecting to a port or local market the areas pro- ducing 80 percent of current agricultural output. If the main product is for export, the link is to a port; if it is for local consumption, the link is to the nearest local market (defined as a town of at least 25,000 people). The difference between the scenarios is in the category of road provided: a single surface treatment road in the base scenario and an all-weather gravel road in the pragmatic scenario. The rural roads included in the assessment of rural connectivity were those needed for trucking consoli- dated agricultural output to the nearest market center or port. To connect areas that provide 80 percent of the value of current agricultural output would require a road network of just under 600,000 km.5 The market-access criterion of rural connectivity was ultimately cho- sen for the estimations reported in this study because it was more univer- sally supported by specialists in transport and rural development. It is no more difficult to calculate the road needs for different standards of mar- ket access than it is for different standards of household proximity to roads. And except at very high levels of connectivity (connecting more than 95 percent of all production areas, for example), the length of the road network required to meet the target was generally less than for meeting comparable RAI standards. But the model retains the capacity to accommodate either method. Urban connectivity. The basic standard of urban connectivity was taken to be a grid of all-weather roads suitable for buses. The grid had to be sufficiently dense that residents would not have to walk great dis- tances to reach the nearest road. The Millennium Cities Database, com- piled by the International Union of Public Transport in 1998, gives data for 100 cities around the world. It shows that, on average, there were about 300 meters of paved road per 1,000 people in the 15 cities ranked lowest on the basis of road density, and about 500 meters per 1,000 peo- ple, on average, for the 20 lowest-ranked cities. Spending to Improve Connectivity 281 Given that the connectivity associated with these figures varies with population density, a maximum walking distance condition was also included in the standard. Thus, for the base scenario, the urban connec- tivity standard was 300 meters of road per 1,000 people and a maximum walking distance of 500 meters, while in the pragmatic scenario, the stan- dard was 200 meters of road per 1,000 people and a maximum walking distance of 1,000 meters. In both cases, the minimum paved road density could be no less than 150 meters per 1,000 people. The connectivity stan- dards were then applied to the total projected urban population of each country in 2015. The urban connectivity standards do not take into account the need for urban mass transit systems such as suburban railways, light-rail rapid transit, subways, or bus rapid transit. Although there is adequate infor- mation on existing systems, no criteria for provision of new systems were included in the standards. Nor do they take into account the addi- tional road capacity that may be needed for private cars.6 The exclusion of requirements for urban mass transit and added road capacity is another respect in which the model's estimates of spending needs may be understated. Category of Infrastructure The base scenario standard for the regional road network was a two-lane, asphalt-paved road, 7.3 meters in width, built with hard shoulders on either side and designed to withstand an 11 tonne single-axle load. National con- nectivity was provided in the model by interurban paved roads of at least one 3.65 meter lane, with shoulders 2.5 meters in width. In practice, politicians would probably argue for a two-lane, 7.3 meter road, as for the regional network. That would increase the costs estimated in this chapter and the financing shortfalls estimated in the next. A compromise that should be explored in further applications of the model would be a combination, determined by traffic volumes, of two-lane asphalt roads, two-lane roads with a slurry seal, and two-lane gravel roads. For rural roads, the base scenario called for a single surface treatment asphalt road. A single surface treatment road was taken as the standard for urban roads in the base scenario (table 7.1). Regional airports were required to have a lighted, paved runway at least 3,000 meters in length, adequate for aircraft used on intercontinen- tal flights, in both the base and pragmatic scenarios. They should have ter- minal space of 20 m2 for each international passenger and 5 m2 for every 1,000 tonnes of air freight. For national connectivity, the runway length 282 Africa's Transport Infrastructure Table 7.1 Definition of the Base Scenario Connectivity Condition of level Connectivity target Infrastructure standard infrastructure Regional Roads Connection of all national Two-lane, paved Good capitals, cities with population of at least 250,000, deep-sea ports, and border crossings Railway Regional network of 39,000 km 20 tonne maximum Good axle load Ports Access to a deep-sea port of One 300 meter berth for Good regional significance with each 0.5 million TEUs appropriate container, general container traffic freight, and bulk freight berths Airports Connection of national capitals At least one runway of Good and cities with population of 3,000 meters at least 500,000 20 m2 terminal space per international passenger National Roads Connection of provincial One-lane, paved Good capitals and cities with population of at least 25,000 Railways Included in regional Same as regional Good Ports Included in regional Same as regional Good Airports Connection of cities with At least one runway of Good population between 100,000 1,524 meters and 500,000 15 m2 terminal space per domestic passenger Rural Roads Market connectivity for One-lane, single surface Good 80 percent of current treatment agricultural production by value Urban Roads No more than 500 meters One-lane, single surface Good walking distance to paved treatment road; 300 meters of paved road per 1,000 people Source: Carruthers, Krishnamani, and Murray 2009. Note: TEU = twenty-foot equivalent unit. was 1,524 meters and the passenger terminal requirement was 15 m2 per domestic passenger. For ports, both the base and the pragmatic standards were that each country should have at least one 300 meter berth for every 0.5 million TEUs (twenty-foot equivalent units) of container freight and for every 5 million tonnes of dry and liquid bulk freight. Spending to Improve Connectivity 283 For regional connectivity in rail transport, it was assumed (based on Bullock [2009]) that rail links should accommodate an axle load of at least 20 tonnes if they were expected to transport more than 5 million net tonnes per kilometer per year. Because the road standards promised to be too expensive for many low-income countries, especially when maintenance was factored in, an improved gravel road with engineered drainage was chosen as the rural standard for the pragmatic scenario, while for urban roads, an improved gravel road with engineered drainage was adopted (table 7.2). Condition of Infrastructure The scenarios also differed in the condition assumed. For the base sce- nario it was assumed that all road and rail infrastructure should be in good condition. Maintenance cost models such as the Highway Development and Management Model (HDM4) attempt to identify the road condition standards and related maintenance expenditures that will keep the total costs of building, maintaining, and using roads to a minimum.7 The base scenario reflects the levels of routine and periodic maintenance implicit in those standards. The lower standards assumed in the pragmatic scenario would reduce overall costs in the 10-year period covered in this analysis but Table 7.2 Definition of the Pragmatic Scenario Connectivity Condition of level Connectivity target Infrastructure standard infrastructure Regional Roads As for base scenario One lane, paved Fair Railways As for base scenario 18 tonne maximum axle load Fair Ports As for base scenario As for base scenario Fair Airports As for base scenario As for base scenario Fair National Roads As for base scenario One lane, single surface treatment Fair Railways As for base scenario As for base scenario Fair Ports As for base scenario As for base scenario Fair Airports As for base scenario As for base scenario Fair Rural Roads As for base scenario One-lane, single surface treatment Fair Urban Roads No more than 1 km One-lane, improved gravel Fair walking distance to road 300 meters of road per 1,000 people Source: Carruthers, Krishnamani, and Murray 2009. 284 Africa's Transport Infrastructure not in the longer term taken into account in the full analyses of the cost-optimization model. While it is assumed in both scenarios that existing transport infrastructure will be initially improved to the level prescribed in the standard, it is also understood that when financial resources are severely constrained, the roads agency may be justified in deferring some periodic maintenance. The ports database was used to gauge the extent to which the iden- tified needs could be met by improvements in the condition of the existing facilities (as well as by extending their lengths or converting no-longer-needed general cargo berths into container berths). The 54 percent of airport runways not already in good condition were assumed to be brought into good condition for both the base and prag- matic scenarios. Applying the Model Applying the model involves three main activities: (i) specifying one or more connectivity scenarios for which spending requirements are to be calculated, (ii) calculating the infrastructure needed to satisfy those requirements in physical terms, and (iii) determining the cost of that infrastructure. Specifying Scenarios Scenarios and their relationship with connectivity targets, infrastructure categories, and condition standards were discussed in the previous sec- tion. Recall that two scenarios were defined: (i) a base scenario that reflects the situation in developed countries and the stated aspirations of many developing countries, and (ii) a pragmatic alternative that reflects the budgetary constraints of many low-income countries--and all low- income, fragile countries. The model is designed to accommodate any other scenario that might be useful. Identifying the Physical Infrastructure Needed to Achieve Connectivity Targets Estimating the network needs for regional and national connectivity was relatively straightforward once the population, GDP, and transport infra- structure databases were established. For urban connectivity, the standard was defined in such a way that estimating the roads needed to satisfy the standard in each country was also straightforward--depending only on the urban population and population density. Spending to Improve Connectivity 285 Rural connectivity presented a more complex challenge. As noted earlier, a standard based on connectivity of agricultural output was adopted. Connectivity needs were calculated using a GIS database of the value of the current agricultural production of each square kilometer of land, as well as the potential value of the same land if it were used to produce the most valu- able crop feasible. Working from this database, the rural road network that would be needed to connect each agricultural zone to the already specified regional and national road networks was assessed, until cumulative values of 20 percent, 40 percent, 60 percent, and 80 percent of the total value of cur- rent national agricultural output were connected. Connectivity was assessed to the nearest significant town (25,000 population or more) or to the near- est port, depending on whether the most valuable crop currently planted in a given zone was destined for local markets or international markets. The rural road network needed to connect zones that provided the same cumulative values of potential agricultural output to the regional and national road networks was similarly assessed. Determining the size of such a network was difficult, however, because in many zones the most valuable agricultural crop--coffee, for example--might have a high export value and a low local value. In other words, more zones would be connected to ports if the criterion of potential value were used. Therefore, the costs of truck- ing output from each zone to the nearest port was subtracted from the value of the highest export crop, and the costs of trucking output from the zone to the nearest significant town was subtracted from the value of the highest locally consumed crop. The connectivity requirement was based on the highest resulting value. In this way, many remote areas that would have been connected to a port using the unadjusted criterion of greatest potential output were connected to the nearest significant town instead. In the next stage of the model, the macroeconomic and GIS databases were applied, together with the scenario definitions, to estimate how much infrastructure was needed to meet the connectivity targets. The resulting estimates were compared with the infrastructure now available in table 7.3, which lists infrastructure elements by total length and quan- tity as well as by category and condition. The GIS data on city, road links, and airport runway locations were used in determining how much of the current infrastructure could be employed to meet specified targets. This exercise revealed that many existing infrastructure facilities were not located where they were most needed. So, although it might appear that more than enough infrastruc- ture is available to meet connectivity needs, much of it is unusable. 286 Africa's Transport Infrastructure Table 7.3 Transport Infrastructure: Current, Base Scenario, and Pragmatic Scenario Transport infrastructure Current Base scenario Pragmatic scenario Regional roads 102,819 km 102,819 km 102,819 km % two-lane 76 100 100 % one-lane 1 0 0 % other 23 0 0 % good condition 32 100 50 National roads 143,531 km 143,531 km 143,531 km % two-lane 44 44 44 % one-lane 3 56 56 % other 54 0 0 % good condition 28 100 50 Rural social criterion n.a. 1,187,050 km 850,450 km Corresponding RAI 29% 75% 50% Local access market criterion n.a. 599,981 km 599,981 km Corresponding RAI n.a. 42% 42% Urban paved roads 39,700 km 111,309 km 49,391 % good condition 15 100 50 Railways, km Length in operation 55,000 km 29,502 km 29,502 km % good condition 38 100 100 Ports Container berths 80 255 255 General cargo berths 159 120 120 Bulk freight berths 105 75 75 % good condition 45 100 100 Airport runways 1,524 to 3,000 meters, number 377 142 142 More than 3,000 meters, number 53 56 56 % good condition 46 100 100 Airport terminals Passenger space, m2 296,500 405,700 405,700 Freight space, m2 57,000 107,000 107,000 Sources: Carruthers, Krishnamani, and Murray 2009 (for base and pragmatic scenarios); CIA 2007; Bofinger 2009; Bullock 2009 (for current). Note: n.a. = not applicable. Meeting the regional connectivity requirements. Roads. Meeting the regional connectivity standard in both the base and pragmatic scenarios would require a network of approximately 103,000 km of two-lane paved roads. The current regional road network includes about 78,000 km of roads built to that standard, and less than 1,000 km of one-lane paved roads that could be upgraded to two lanes (table 7.3). To meet the targets, therefore, some 20,000 km of gravel roads that make up part of Spending to Improve Connectivity 287 the current regional network would have to be upgraded, together with about 6,500 km of dirt roads and 3,500 km of dirt tracks. Only 32 per- cent of the network is in good condition, a reflection of the very poor con- dition of unpaved roads. Because the proposed Trans-African Highway system is counted toward the regional connectivity standard, the model took into account all of the system's links, including segments planned but not yet built. The investment requirements presented here include the 52,450 km needed to complete and fully pave that system--and more. The Trans-African Highway was designed to connect capital cities to a common road net- work, whereas the standard proposed here for regional connectivity also calls for connecting large cities, ports, and airports. The length of road needed to meet this regional connectivity standard, therefore, is about double that of the Trans-African Highway. The construction specifica- tions, however, are the same. Airports. To achieve the regional airport connectivity standards would require a total of 56 runways of 3,000 meters and an additional 405,700 m2 of passenger terminal space and 107,000 m2 of freight ter- minal space (table 7.4). Although the number of total runways needed is only 3 more than the number currently available, 8 of the latter are in cities smaller than those specified in the connectivity standard, so 11 additional runways will be needed. Several of these can be provided by extending existing shorter runways. Ports. A total of 255 container berths, 75 bulk freight berths, and 120 general freight berths would be required to service the interna- tional maritime trade of the African countries. The number of general freight berths needed will be substantially fewer than the current 159, as most general freight will be containerized within the next 10 years. Table 7.4 Regional Airports: Current Characteristics vs. Characteristics Needed to Meet Connectivity Target Target Regional airports Current Base scenario Pragmatic scenario Airport runways Longer than 3,000 meters, number 53 56 56 % in good condition 46 100 100 Airport terminals Passenger space, m2 296,500 405,700 405,700 Freight space, m2 57,000 107,000 107,000 Source: Carruthers, Krishnamani, and Murray 2009. 288 Africa's Transport Infrastructure The projected (solid) bulk freight berths do not include those in facili- ties for specific bulk products (such as grain terminals adjacent to ports for general cargo and containers). Railways. Bullock (2009) estimates that there are about 65,000 km of railways in Africa, of which about 56,000 km are in operation; 31 percent of those in operation are in South Africa. Although Bullock does not spec- ify their condition, he estimates the length of national networks that could be economically viable. Low-volume lines (those that carry less than 1 million net tonnes of freight per year) are unlikely to merit full rehabilitation, and lines carrying bulk traffic will normally generate greater (and more certain) benefits than those carrying general traffic. Applying these criteria to national rail networks, Bullock estimated that only about 29,500 km of line would have enough traffic to justify the expenditure needed to keep them in operation. This is the figure that was used here in computing the cost of upgrading to the standard of 20 tonne axle loads. Meeting the national connectivity requirements. Roads. Only about 47 percent of the 143,500 km of roads needed to meet the national road connectivity standard are presently paved. Meeting the standard would require upgrading gravel roads and earth roads in approxi- mately equal proportions. The share of national roads in good condi- tion is just under 30 percent, significantly less than the share of regional roads in good condition. This is because more of the national roads are unpaved, and unpaved roads tend to be in worse condition than paved roads. Airports. A total of 142 runways of between 1,524 and 3,000 meters are needed to meet the national airport standard. While this number is much smaller than the 377 runways already found in Africa, many of those are in cities smaller than the ones specified in the standard. Others appear to be for military use only, and a few will need to be extended. For technical reasons, all of the terminal space for domestic air passengers is included in the total required for regional connectivity. It is estimated that about 50 percent of the 400,000 m2 of additional passenger terminal space needed to meet the regional connectivity targets would be for domestic passengers. No allowance was made for freight terminal space, which is likely to increase in the future. The national connectivity targets for ports and railways are the same as the regional targets. Spending to Improve Connectivity 289 Meeting the rural connectivity requirements. A very long network of roads would be required to provide the degree of social accessibility cor- responding to an RAI value of 100 percent (table 7.5). The network length required to meet the standard of market access for 80 percent of the value of agricultural output is a good deal lower. For provision of market access for 80 percent of current agricultural output (by value), just under 600,000 km of road would be required-- possibly 50 percent more than the current total, though exact figures for the current total of relevant roads are not available. But because many current roads are either poorly located or in bad condition, substantial upgrades and new construction are necessary. For connection of 80 percent of potential agricultural output, up to six times more new roads would be needed (table 7.6). Meeting the urban connectivity requirements. In urban areas, about 111,000 km of paved roads are required to meet the connectivity target specified in the base scenario for the projected urban population (table 7.7). The pragmatic scenario would require almost 50,000 km. No African country now has a network of paved urban roads large enough to satisfy the base scenario; only two (Lesotho and Namibia) fulfill the requirements of the pragmatic scenario. Presently there are only about 40,000 km of paved roads at least one lane wide in the urban areas of Africa, so that about 71,000 km of additional urban paved roads would be required for the base scenario and about 9,000 km for the pragmatic scenario. The main source of these additional paved roads would be existing gravel and earth roads. But even making use of these, to reach the base scenario's urban connectivity standard would require more than 30,000 km of new Table 7.5 Road Lengths Needed to Reach Rural Connectivity Standards Road length Resulting RAI value Road network (km) (%) Network needed to achieve 100% RAI 1,473,602 100 Network needed to achieve 75% RAI 1,187,050 75 Network needed to achieve 50% RAI 850,450 50 Network needed to connect 80% of current agricultural output 599,981 40 Network needed to connect 80% of potential agricultural output 912,487 55 Source: Carruthers, Krishnamani, and Murray 2009. 290 Africa's Transport Infrastructure Table 7.6 Sources of Road Upgrades to Achieve Rural Connectivity Targets To connect 80% of To connect 80% of current potential agricultural Road network agricultural output (km) output (km) Total rural roads needed to satisfy connectivity criteria 599,981 912,487 From regional + national 243,356 242,309 From other classified paved roads 153,998 163,617 From classified unpaved roads 166,984 308,668 From unclassified paved roads 3,600 24,188 From unclassified unpaved roads 9,163 37,971 New roads or upgraded tracks 22,879 135,733 Rural Accessibility Index corresponding to this network (%) 40.0 55.2 Precentage of new roads or upgraded tracks in rural total 3.8 14.8 Source: Carruthers, Krishnamani, and Murray 2009. Table 7.7 Types of Road Counted toward Urban Connectivity Targets Target Road type Base scenario (km) Pragmatic scenario (km) Total paved roads needed to meet 110,880 49,199 connectivity target Existing paved roads 34,894 29,802 Existing gravel roads 30,941 9,238 Existing dirt roads 12,466 3,544 New paved roads 32,579 6,615 Sources: http://www.geohive.com/default1.aspx; Carruthers, Krishnamani, and Murray 2009. roads, with fewer than 7,000 km needed for the pragmatic scenario. Because not all of the existing roads are in the right place, however, the actual need for new roads is likely to be somewhat greater than these figures suggest. Estimating the Cost of Meeting the Connectivity Requirements under Both Scenarios Once infrastructure needs and the means of meeting them are known, the cost matrix can be applied to estimate the cost of meeting connectivity targets. Spending to Improve Connectivity 291 Previous assessments of spending requirements in the transport sector (Fay and Yepes 2003; Estache and Yepes 2004; Chatterton and Puerto 2005) have focused on the costs of lengthening road and rail networks and then maintaining them. The target length of those networks usually has been derived from a simple econometric model or a benchmarking approach that uses industrial economies as comparators. The difference between the target and the existing network is then multiplied by the unit cost of new infrastructure to derive the cost necessary to bring the network up to the target level. Estimates of maintenance costs typically have been even more simplistic and are usually given as a percentage of the replacement value of the network. These simplified approaches reflect the paucity of data on quantities and qualities of transport infrastructure. The model employed in this study makes use of many data sources not used by earlier investigators, several of which were compiled specifically for the purpose. These expanded and more reliable data allow better estimates of spending needs, despite some remaining gaps. Four different types of infrastructure costs were estimated: · The cost of improving the condition of current transport infrastructure to the standard defined in each scenario · The cost of upgrading the standard of existing transport infrastructure to that defined in each scenario (for example, widening existing roads or upgrading their surface; extending existing airport runways, passen- ger terminals, and port berths; and increasing the permissible axle load of railways) · The cost of extending existing networks--such as regional, national, rural, and urban roads--and increasing the number of infrastructure assets, such as port container berths, to supplement existing transport infrastructure and so reach the scenario targets · And finally, the largest category, the cost of maintaining networks and assets--in their improved, upgraded, or expanded form--in the condi- tion defined in each scenario Each type of cost is dealt with in turn. The cost of improving infrastructure conditions. Where available, data on current conditions (good, fair, or poor) were used; where such esti- mates could not be found, assumptions were made based on the propor- tion in each condition in comparable countries. To estimate the cost of 292 Africa's Transport Infrastructure bringing infrastructure in fair or poor condition up to good condition, the quantities of such infrastructure were multiplied by the unit cost of improvement--a one-time cost that can be incurred at any time. It was assumed that this cost could be spread equally over 10 years. The cost of upgrading infrastructure standards. Infrastructure is catego- rized by its capacity. The essential question is whether a piece of infra- structure has the capacity to meet the connectivity demands made upon it. (For example, airports in many large cities have runways that are too short for the midsize aircraft--such as the Airbus A320 or Boeing 737-- used on domestic trunk routes.) Hence, the infrastructure standards appropriate for each mode and market are specified. In some cases, these specifications were based on international standards, such as those of the International Civil Aviation Organization. In other cases, they were based on common engineering standards or on authors' judgment and experi- ence. The costs of making the necessary upgrades were based generally on engineering estimates. The costs of upgrading infrastructure in poor or fair condition were estimated in two stages: first, improving its condition to "good" at its pres- ent standard, and then upgrading it to the next standard. Where infra- structure was already of a higher standard than deemed necessary for its purpose, it was assumed that it would be maintained in good condition at its current standard. The cost of extending network length and increasing facilities. A combi- nation of methods was used to estimate the optimal or desired extent of transport infrastructure networks or assets. For each type of road, the net- work length needed to meet connectivity targets was compared against cur- rent length, estimated using GIS data. Optimum airport runway length was estimated for each city, also using GIS data. For port berths, the require- ment was based on the port connectivity standard, applied to the output of a macroeconomic model that included, for each country, projections of GDP, imports and exports as a share of GDP, and the average value per tonne of freight (or per TEU for containerized trade). The resulting target quantities were compared with present levels, regardless of condition. The costs of extending networks or increasing the number of overall assets were then estimated. (The costs of bridges or tunnels were not included in either urban or interurban road-cost estimates.) For urban roads, neither the capacity needs of private cars (greater than those required for buses) nor the costs of urban mass transit systems were taken into account. Spending to Improve Connectivity 293 For interurban roads and railways, the methods used resemble those of previous studies. But for improving or maintaining other modes (rural and urban roads, airports, and ports), the cost estimates presented here are difficult to compare with those of other studies. To facilitate compar- isons, however, this study disaggregates final cost estimates by transport mode and type of expenditure. The cost of maintaining infrastructure. Most previous assessments of transport infrastructure costs made some attempt to include maintenance costs by adding a fixed amount (often 3 percent) of the replacement cost of the infrastructure. For most transport infrastructure, two types of main- tenance were considered here--annual and periodic. In the model, both types are taken into account, but periodic costs are converted to annual sums. The resulting annual average costs of maintenance are specific to each country, although the unit costs are the same. This is because the better the current condition of an asset, the less needs to be spent on peri- odic maintenance. The annual cost of maintaining a facility in good con- dition is taken to be the same for any given type of infrastructure across all countries. While in some cases the estimates of maintenance costs were close to 3 percent of asset replacement costs, in general the esti- mates were higher than those of previous assessments, with a wide varia- tion about the average. Applying unit costs to the identified needs. Once all the necessary infra- structure operations are known (for example, the upgrading of earth roads or the construction of new facilities), the results are fed through a matrix of unit costs for building, upgrading, and maintaining each type of transport infrastructure at the levels defined in each scenario (Carruthers, Krishnamani, and Murray 2009). Data on unit costs used in the matrix were derived from fieldwork in a sample of southern African countries. The total expenditure for each scenario that results from these calcula- tions is the final output of the model. Outputs of the Model The outputs of the model can be expressed in global terms, or disaggre- gated by country, mode, or type of expenditure. Costs are provided as totals for a 10-year period, as annual averages for each of the 10 years, and as a percentage of annual GDP. A Web-based version of the model is available at https://www.infrastructureafrica.org. Each output offers 294 Africa's Transport Infrastructure useful insights into the fundamental policy issues surrounding transport infrastructure provision. Total Spending Needs in the Base Scenario Considering the total expenditure needs under different scenarios high- lights the connections between needs, affordability, and fundamental aspirations. At about $12.7 billion, the annual cost of meeting the connectivity targets for transport in the base scenario would average slightly less than 2 percent of GDP each year for the 10 years between 2006 and 2015 (table 7.8). Maintenance of improved, upgraded, and expanded infra- structure would require almost 40 percent of the total expenditure, with upgrading requiring a further 30 percent, improvement about 20 percent, and expansion less than 15 percent. These allocations are substantially different from those made today, when construction and expansion take up the largest shares. The total expenditure estimates listed in table 7.8 have been adjusted to reflect the impressive transport investments made by the South African government in preparation for the 2010 football World Cup. The 2010 World Cup transport projects included the massive expansion of public transport and road infrastructure, rail upgrades, development of intermodal facilities, bus rapid transit systems, inner-city mobility systems, call-center systems, airport-city links, freight services, and passenger safety and intelli- gent transport systems. The flagship projects--which by themselves totaled over $3 billion--included the Khulani corridor in the Eastern Cape (R 321 million), the N1 and N2 Toll Highway (R 5 billion), the Sani Pass road upgrade on the border between South Africa and Lesotho (R 200 million), and the Gauteng Freeway Improvement Scheme (R 23 billion). Another massive project is the expansion of airport facilities undertaken by the Airports Company South Africa, with an estimated cost of $1.5 billion. Putting the Numbers into Perspective: Affordability The 1.9 percent of GDP needed to achieve the connectivity standards of the base scenario is about the same as what the countries of the European Union invest in their transport infrastructure. But the European Union average is for countries that have been investing in their transport infra- structure for centuries, with few interruptions or periods of significant deferred maintenance. Perhaps a more realistic comparison is with the middle-income countries that a half-century ago were at a stage of devel- opment similar to that of Africa today. Brazil, Japan, and Republic of Table 7.8 Average Annual Investment Needs: Base Scenario, 2006­15 Expenditure purpose Improving Upgrading Expanding Maintaining condition standard capacity infrastructure As percentage of Infrastructure type US$ (millions) Total GDP Total expenditure Regional roads 519 1,083 172 905 2,678 0.4 21 National roads 451 1,237 174 1,015 2,877 0.5 23 Rural roads 812 397 63 1,205 2,477 0.4 19 Urban roads 275 566 684 622 2,147 0.3 17 Airports 40 23 84 710 856 0.1 7 Ports 203 70 449 153 876 0.1 7 Railways 128 265 100 282 776 0.1 6 Total for all types (without South Africa's World Cup investment needs) 2,429 3,642 1,726 4,891 12,689 1.9 100 As % of GDP 0.4 0.6 0.3 0.8 2.0 As % of total 19 29 14 38 100 Total for all types (with South Africa's World Cup investment needs) 2,429 3,642 3,569 9,553 19,193 2.9 100 Source: Carruthers, Krishnamani, and Murray 2009. 295 296 Africa's Transport Infrastructure Korea--and even the Federal Republic of Germany in its post­World War II reconstruction phase--invested much higher proportions of GDP in their transport infrastructure. Some invested between 5 and 8 percent of GDP during the 1950s and into the 1960s--their period of highest eco- nomic growth (Korea's transport development was later than for the other countries). Since the 1990s, China has been investing more than 6 percent of its GDP in transport infrastructure, largely as part of its strategy to advance the relative economic growth of its inland provinces. The amount that a country can afford to spend on transport depends on much more than sectoral needs alone. But in few instances have countries invested more than 5 percent of their GDP for a decade or more, while those countries that have invested less than 1 percent of their GDP have found such low levels to be unsustainable and have later increased invest- ment. For example, China allowed its transport investment to fall to 1 per- cent of GDP for almost a decade before raising the rate to more than 6 percent in the 1990s (World Bank 1998). Since the per capita needs for transport investment tend to increase less quickly than GDP per capita, countries with a higher GDP per capita usually need to invest a smaller share than those with a lower GDP. On the other hand, countries that have invested less in the past need to invest more now to reach similar standards of connectivity. Other factors, such as prolonged conflicts, also influence required expenditures. During civil war, for example, not only are roads and bridges destroyed as strategic targets, but infrastructure maintenance becomes unsustainable. Civil war thus increases the cost of achieving con- nectivity standards even more than a history of sustained underinvestment. Total Spending Needs in the Pragmatic Scenario The objective of the pragmatic scenario is to reduce the spending require- ments of the low-income countries (including the fragile countries) and to a lesser extent of the resource-rich countries. And the pragmatic scenario (table 7.9) does entail less spending: the average difference across Africa between the base and pragmatic scenarios is 40 percent (tables 7.8 and 7.9). The base scenario appears sustainable for the region's few middle- income countries--whether as an absolute amount or as a percentage of GDP. But for most other country groups, meeting the lower standards of the pragmatic scenario seems to be a feasible interim solution, with the base scenario remaining as a long-term objective. For a small number of countries, achieving even the pragmatic standards is not feasible in the short to medium term without massive external aid; such countries may need to consider reducing their connectivity objectives even further. Table 7.9 Average Annual Investment Needs: Pragmatic Scenario Investment purpose Improve Upgrade Expand Maintain condition standard capacity infrastructure As percentage of Infrastructure type US$ (millions) Total GDP Total expenditure Regional roads 157 464 103 858 1,582 0.2 20.9 National roads 132 928 131 1,012 2,202 0.3 29.1 Rural roads 99 98 63 1,363 1,624 0.3 21.5 Urban roads 275 143 119 341 879 0.1 11.6 Airports 43 23 48 152 266 0.0 3.5 Ports 204 70 77 153 504 0.1 6.7 Railways 128 199 100 79 506 0.1 6.7 Total 1,039 1,925 641 3,958 7,563 1.2 100.0 As % of GDP 0.2 0.3 0.1 0.6 1.2 As % of total 13.7 25.5 8.5 52.3 100.0 Source: Carruthers, Krishnamani, and Murray 2009. 297 298 Africa's Transport Infrastructure As noted, for Africa, excluding South Africa, the average annual cost of the pragmatic scenario is only 60 percent of the base scenario's cost ($7.6 billion vs. $12.7 billion). But the reduction is not equally distrib- uted across expenditure types; for example, the difference between the scenarios is small for both maintenance and new construction but very large for infrastructure upgrades and improvements. Of the 13 countries that would need to spend more than 5 percent of their GDP to achieve the base scenario standards, 9 are designated as low- income, fragile states; two, Madagascar and Mauritania, are nonfragile, low-income countries; and two, the Central African Republic and Chad, are designated as resource-rich states (figure 7.2). It is these 13 countries that would benefit most from reducing their transport sector expecta- tions to standards lower than those of the base scenario. By shifting from the base to the pragmatic scenario, the average expendi- ture need for this group would fall from an average of more than 11 percent of GDP to less than 7 percent. Guinea-Bissau would enjoy the largest Figure 7.2 Scenario Costs in Countries Where the Base Scenario Would Cost More than 5 Percent of GDP Liberia Congo, Dem. Rep. Central African Republic Guinea-Bissau Sierra Leone Gambia, The country Guinea Madagascar Togo Zimbabwe Mauritania Chad Mozambique 0 5 10 15 20 25 30 35 40 % of GDP pragmatic % of GDP base Source: Carruthers, Krishnamani, and Murray 2009. Spending to Improve Connectivity 299 percentage reduction in its needs (about 60 percent), while The Gambia would have the smallest gain (saving just 15 percent on its spending needs); the average reduction for these countries would be 42 percent. Liberia would have the largest reduction in absolute terms, from more than 37 per- cent of its GDP to less than 19 percent. Mixing the scenarios--applying dif- ferent scenarios to different states--is another possibility (box 7.1). Box 7.1 Mixing Scenarios The variety of challenges facing the countries of Africa requires a variety of responses. Although for reasons of space this study does not explore a wide range of alternatives, the model is capable of doing so if required. As an example, the pragmatic scenario could be applied to the 13 countries that would have to spend more than 5 percent of their GDP to achieve the base scenario, while the base scenario could be applied to the others. Because 9 of those 13 countries are low-income, fragile states, it would also be possible to isolate them as candidates for the pragmatic scenario. The outcomes of these two options would be rather different in terms of the total spending requirements for all countries concerned. If the pragmatic scenario were applied to all the low-income, fragile states, but not to other low-income states, the total expenditure would be about $11.3 billion per year, about 10 percent less than the base scenario. If instead it were applied to all countries where the base scenario would require investment of more than 5 percent of GDP, the total spending requirement would be about $10.7 billion, or 15 percent less than that of the base scenario (see table B7.1). There are advantages and disadvantages to either method of blending sce- narios, and the choice between them will depend on the objective being sought. Table B7.1 Comparing Blends of the Base and Pragmatic Scenarios Spending requirement % of base Scenario US$ (billions) % of GDP scenario Base 12.689 2.0 100 Pragmatic 7.563 1.2 60 Pragmatic for all low-income, fragile states 11.331 1.8 90 Pragmatic where spending requirement in base scenario is at least 5% of GDP 10.688 1.7 85 Source: Carruthers, Krishnamani, and Murray 2009. 300 Africa's Transport Infrastructure Spending Needs by Country and Country Group Low-income, fragile states would need to spend more than 8 percent of their GDP to achieve the connectivity standards of the base scenario, while the other low-income countries would need to spend just 3.4 percent of GDP--still a high percentage (table 7.10). Meanwhile, resource-rich countries would have to spend 1.7 percent of GDP, and the middle- income countries just 0.7 percent of their much higher GDP. Spending needs as a share of GDP vary more widely when individual countries are considered, ranging from a high of more than 37.0 percent of GDP for Liberia to a low of 0.5 percent for South Africa. The prag- matic scenario, though aimed at countries for which the base scenario is unaffordable, would reduce costs across all groups of countries. Considering the region as a whole, the resource-rich countries and other low-income countries would each require about 30 percent of all spending under the base scenario, with the low-income, fragile states taking 25 percent and the middle-income countries the remain- ing 15 percent. Those shares are not in proportion to their share of the region's total GDP or population. The low-income, fragile states have only 5 percent of the region's total GDP and 20 percent of the total population, but they would need 25 percent of the total expenditure to meet the connectivity standards of the base scenario. The other country groups would require a share of the spending envelope that falls between their share of GDP and of population. The shares of total spending needs absorbed by the various country groups change when shifting from the base to the pragmatic scenario, Table 7.10 Transport Infrastructure Expenditure as Percentage of GDP, by Country Group and Scenario Base scenario Pragmatic scenario Country group % of GDP % of total investment % of GDP % of total investment Low-income, fragile 8.2 30 4.8 24 Low-income, nonfragile 3.4 25 2.2 32 Resource-rich 1.7 30 1.0 31 Middle-income 0.7 15 0.4 13 Average of all African countries 2.0 n.a. 1.2 n.a. Source: Carruthers, Krishnamani, and Murray 2009. Note: n.a. = not applicable. Spending to Improve Connectivity 301 although the share required by the middle-income countries changes by just two percentage points and that required by the resource-rich coun- tries by one point. For the low-income, fragile states, the share drops from 30 percent in the base scenario to 24 percent in the pragmatic, while the share of the other low-income countries increases from 25 percent to 32 percent. These fluctuations indicate that the reduced standards of the pragmatic scenario have the desired impact of reducing costs the most in the vulnerable low-income, fragile countries, and the least in the less-vulnerable countries. There are at least three ways of normalizing expenditure: per dollar of GDP, per capita, and per unit of land area (table 7.11). Of the 10 countries with the greatest spending needs per dollar of GDP, 9 are categorized as low-income, fragile states. By contrast, low- income, fragile countries do not figure prominently in the rankings of investment per capita or by land area. Liberia--with by far the highest expenditure needs per dollar of GDP--ranks only 29th in terms of expenditure needs per capita, but is 9th in relation to land area. Of the 10 highest-ranked countries by expenditure needs per dollar of GDP, 7 also appear in the top 10 using one of the other two criteria. Understandably, sparsely populated countries tend to rank high in terms of spending needs per capita, and countries with a small land area rank high in terms of investment cost per square kilometer--but there are significant exceptions to these generalizations. Although all of the 10 highest ranked countries by expenditure needs per capita have popu- lations of less than 5 million, the 12th-ranked country, South Africa, is ranked 4th in terms of total population. Of the 10 highest-ranked coun- tries in terms of expenditure needs per square kilometer of land area, 9 have relatively small areas (being ranked in the last 10 in terms of size). The exception here is the Democratic Republic of Congo, which is ranked 5th in terms of expenditure needs per square kilometer and 21st in terms of total land area. Spending Needs by Purpose of Spending In both the base and the pragmatic scenarios, infrastructure maintenance requires the largest share of expenditure--about 39 percent in the base scenario (figure 7.3) and more than 53 percent in the pragmatic. Spending for maintenance also drops less than spending for other purposes in the shift from the base to the pragmatic scenario. While spending for mainte- nance falls from more than $4.9 billion per year in the base scenario to less than $4 billion in the pragmatic, a drop of 20 percent, spending to expand 302 Table 7.11 Highest Transport Infrastructure Spending Needs by Country, Relative to GDP, Population, and Land Area (base scenario) Ranked by percentage of GDP Ranked by US$ per capita Ranked by US$/km2 Rank Country Percentage Rank Country US$/pop Rank Country US$/km2 1 Liberia 37.25 1 Equatorial Guinea 148.43 1 Mauritius 37,469 2 Congo, Dem. Rep. 20.72 2 Gabon 125.62 2 Guinea 8,940 3 Central African Republic 18.55 3 Mauritania 95.15 3 Cape Verde 4,621 4 Guinea-Bissau 12.72 4 Mali 78.83 4 Swaziland 4,472 5 Sierra Leone 9.92 5 Namibia 75.91 5 Congo, Dem. Rep. 4,309 6 Gambia, The 8.94 6 Central African Republic 59.62 6 Gambia, The 4,121 7 Guinea 7.69 7 Botswana 59.21 7 Equatorial Guinea 2,646 8 Madagascar 7.48 8 Congo, Rep. 44.17 8 Togo 2,494 9 Togo 6.58 9 Djibouti 41.84 9 Liberia 1,773 10 Zimbabwe 6.38 10 Cape Verde 37.65 10 Sierra Leone 1,682 Source: Carruthers, Krishnamani, and Murray 2009. Spending to Improve Connectivity 303 Figure 7.3 Comparison of Scenarios by Type of Spending 6,000 5,000 annual investment (in US$ billions) 4,000 3,000 2,000 1,000 0 maintain upgrade improve new scenario base pragmatic Source: Carruthers, Krishnamani, and Murray 2009. capacity falls by more than 63 percent, to improve condition by 57 percent, and to upgrade category by 47 percent. The ratio of spending for mainte- nance to total spending is much greater in the model presented here than the actual ratio achieved in recent decades (most countries allocated to maintenance an average of between 25 percent and 33 percent of their total infrastructure spending). The higher figure here reflects the position taken by many lending institutions and sector professionals that it is often more productive and efficient to maintain existing infrastructure than to build new. The balance of spending by purpose also varies by country groups. For example, middle-income countries have generally less need to upgrade their infrastructure and, hence, need to spend proportionately more on maintenance (table 7.12). Details of the spending needs by purpose for the pragmatic scenario are shown in appendix 6. Spending Needs by Type of Connectivity Regional, national, and rural roads each account for between 20 and 30 percent of overall spending needs in both scenarios (figure 7.4). Urban roads absorb a share of almost 17 percent of spending in the base scenario but less than 12 percent in the pragmatic. Airports, ports, and railways each receive between 6 and 7 percent of investment in the base scenario. These shares remain at more than 6 percent for ports and railways in the pragmatic scenario, while the airport share drops below 4 percent. 304 Africa's Transport Infrastructure Table 7.12 Spending Needs by Purpose for Each Country Group (base scenario) Improve Upgrade Expand Maintain infrastructure infrastructure infrastructure infrastructure (%) (%) (%) (%) Low-income, fragile 20 36 14 30 Low-income, other 17 32 8 43 Middle-income 21 11 21 48 Resource-rich 19 28 15 37 All African countries 19 29 14 39 Source: Carruthers, Krishnamani, Murray, and Pushak 2009. Note: Rows may not total 100 percent due to rounding. Figure 7.4 Comparison of Scenarios by Mode of Transport 4,000 annual spending 3,000 (US$ billions) 2,000 1,000 0 s al l n r rt il na es ai ra ba po on cc tio ur gi ta na re ke ar m mode base pragmatic Source: Carruthers, Krishnamani, and Murray 2009. Overall spending on airports is almost 70 percent less in the base than in the pragmatic scenario. Within either of the scenarios, the allocation of investment needs by mode varies substantially between country groups. For example, the middle-income countries have proportionately less need for expenditure on regional, national, and rural roads, but greater needs for ports, air- ports, and railways (table 7.13). As with the purpose of spending, the allocation of spending across modes of connectivity in both scenarios differs from actual experience in recent years. Spending for regional and national roads under the scenar- ios proposed here would be lower than at present, while spending for Spending to Improve Connectivity 305 Table 7.13 Percentage Allocation of Investment Needs by Country Group and Transport Mode (base scenario) Low-income, Low-income, Middle- All African Connectivity/mode Resource-rich fragile other income countries Regional roads 23 24 20 14 21 National roads 22 23 26 17 23 Rural roads 20 26 22 2 20 Urban roads 18 17 15 19 17 Railways 6 4 6 12 6 Ports 7 2 4 20 7 Airports 4 3 8 16 7 Total 100 100 100 100 100 Share of total (%) 30 25 30 15 100 Source: Carruthers, Krishnamani, and Murray 2009. Note: Columns may not total 100 percent due to rounding. rural and urban roads would increase substantially, as rural roads have been neglected, except in the middle-income countries. The greatest burden on the poorer countries is associated with improving rural connectivity. The costs (as a percentage of GDP) of con- necting different shares of current and potential agricultural production by a paved road with a single surface treatment, as in the base scenario, are illustrated in figure 7.5. The dotted line shows that the average cost of connecting 80 percent of present output, by value, would be about 0.4 percent of GDP, while that of connecting 80 percent of potential output exceeds 0.5 percent. The cost burden of this level of rural market connectivity varies greatly across countries. The solid line in figure 7.5 shows that the region's low- income, fragile states would need to spend more than 2 percent of their GDP to connect 80 percent of current output, by value, and 3.5 percent of GDP to connect the same share of potential output. The most signifi- cant differences in the cost of rural connectivity for African countries taken together and for the subset of low-income, fragile states arise when one attempts to achieve a high degree of connectivity for areas represent- ing rising shares of the value of potential rather than current agricultural output. The emergence of the gap reflects the situation in the middle- income countries, where adding connectivity for 80 percent of the value of potential agricultural output using sealed roads would represent an investment of no more than 0.05 percent of GDP. 306 Africa's Transport Infrastructure Figure 7.5 Rural Connectivity Costs for Different Percentages of Agricultural Production, by Value 4.0 3.5 3.0 share of GDP 2.5 2.0 1.5 1.0 0.5 0.0 20 40 60 80 +20 +40 +60 +80 percentage of agricultural production linked low-income, fragile Africa Source: Carruthers, Krishnamani, and Murray 2009. Note: Shares of GDP required have been calculated for the proportions of output connected marked, intermediate values have been interpreted to give a continuous line. The cost of rural market connectivity for low-income, fragile states can be reduced to a cost of less than 1.5 percent of GDP by using all-weather gravel roads, as specified in the pragmatic scenario (figure 7.6). Insights from the Connectivity Analysis It should be remembered that this chapter describes a modeling exercise, the results of which are critically sensitive to the inputs. A number of ele- ments that could lead to underestimating costs have been noted: · Bridges and some other ancillary investments are excluded from the calculations. · Urban roads that meet the connectivity targets in aggregate may not be suitably located for current development patterns. · Standards may be set too low, particularly for national connectivity by road. · Costs may be too low, particularly for landlocked countries or those with difficult terrain. Spending to Improve Connectivity 307 Figure 7.6 Rural Connectivity Costs for Low-Income, Fragile States with Different Categories of Roads 4.0 3.5 3.0 share of GDP 2.5 2.0 1.5 1.0 0.5 0.0 20 40 60 80 +20 +40 +60 +80 percentage of agricultural production linked SST all weather gravel Source: Carruthers, Krishnamani, and Murray 2009. Note: SST = single surface treatment. Shares of GDP required have been calculated for the proportions of output connected marked, intermediate values have been interpreted to give a continuous line. The implication of this range of possible error in the cost estimations, which is also probably asymmetric, is that some countries are in a signif- icantly worse situation than the financing shortfall estimations of the next chapter suggest. Bearing that caveat in mind, the analysis still strongly suggests the need to reappraise the current allocation of public expenditure on transport infrastructure. For achievement of common connectivity targets--that is, to bring people closer to jobs, services, and markets--a much greater share of investment must be directed to local rural road access and urban roads, and, to a lesser extent, to railways, airports, and seaports. The recent wave of concessioning in the latter areas has helped redress the lack of investment, but many operations cannot earn sufficient revenue from their operations to attract private investment in infrastructure. Public-pri- vate partnering will be needed to close the investment gap. Knowing how much infrastructure is needed to achieve any particular social or economic target is meaningless if, once obtained, the infrastructure cannot be maintained. The analysis shows that almost half of all spending on transport infrastructure will be needed to maintain the roads and other infrastructure that are improved, upgraded, and expanded under the base and pragmatic scenarios. (If the condition of infrastructure is not improved, 308 Africa's Transport Infrastructure even more will have to be spent on maintenance.) Because maintenance standards are lower under the pragmatic scenario, and because infrastruc- ture networks and capacity are not expanded as far, spending for mainte- nance drops sharply in the pragmatic scenario but still dwarfs investment in improvements, upgrades, and extensions of infrastructure. The proportion of transport expenditure allocated to maintenance has never come close to the 40 to 50 percent share shown here to be needed, despite the establishment of road maintenance funds in many of the countries included in this analysis. This persistent underspending on maintenance suggests that the road networks of many countries in the region may become unsustainable--unless a more judicious mix of investments can be found and followed. The Asian Development Bank (ADB 2006) estimated the road net- work size that can be sustained by public sources. After reviewing current network size and quality, as well as actual maintenance spending, the ADB study found a negative relationship between network quality and network replacement value as a share of GDP. When the replacement value exceeds about 40 percent of current GDP, quality declines rapidly. Using this finding, the study concluded that the road network density and standards of several countries in the sample, particularly large countries with low GDP, were greater than their economies could sustain. The fore- going analysis supports that conclusion. Notes 1. This chapter is based on Carruthers, Krishnamani, and Murray (2009). 2. However, the capacity of the facilities may be less than would normally be considered acceptable for main trunk routes in an industrialized country; see the section below on category of infrastructure. 3. The infrastructure elements included in the database are detailed in the indi- vidual country maps appearing in the country annex to Carruthers, Krishnamani, and Murray (2009). 4. In Uganda, no city but the capital has a population exceeding this threshold; in Ghana, there is only one other (Kumasi). Meanwhile, in Nigeria, there are 26 and in South Africa, over 30. 5. The alternative of extending the base scenario to provide connectivity to areas representing 80 percent of the potential value of output (that is, the value of output if the most lucrative possible crops were grown) was also considered but not adopted, as the cost would have been prohibitive for all countries involved. Spending to Improve Connectivity 309 6. This approach is defensible for low-income cities where the level of con- gestion is determined primarily by public transport, freight, and nonmotor- ized traffic. 7. The HDM4 is a computer software system for investigating choices of expen- diture on road transport infrastructure. Originally developed by the World Bank, responsibility for management and further development of the suite of programs was taken over by the World Road Association (PIARC) in 1996. The current version was finalized in June 2005. It is now managed under a service concession contract with HDMGlobal, an international consortium led by the University of Birmingham, England. References ADB (Asian Development Bank). 2006. "Road Asset Management." Technical Assistance Report 5925, Asian Development Bank, Manila. Bofinger, H. C. 2009. "Air Transport: Challenges to Growth." Africa Infrastructure Country Diagnostic Background Paper 16, World Bank, Washington, DC. Bullock, R. 2009. "Railways in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 11, World Bank, Washington, DC. Carruthers, R., K. Krishnamani, and S. Murray. 2009. "Improving Connectivity: Investing in Transport Infrastructure in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 7, World Bank, Washington, DC. Chatterton, I., and O. S. Puerto. 2005. "Estimation of Infrastructure Investment Needs in the South Asia Region." Unpublished paper, World Bank, Washington, DC. CIA (Central Intelligence Agency). 2007. The World Factbook. https://www.cia.gov/ library/publications/the-world-factbook/. Demographia. 2009. Demographia World Urban Areas and Population Projections. 5th ed. Belleville, IL: Wendell Cox Consultancy. Estache, A., and T. Yepes. 2004. "Assessing Africa's Infrastructure Needs." Unpublished paper, World Bank, Washington, DC. Fay, M., and T. Yepes. 2003. "Investing in Infrastructure: What Is Needed from 2000 to 2010?" Policy Research Working Paper 3102, World Bank, Washington, DC. Gwilliam, K., V. Foster, R. Archondo-Callao, C. Briceño-Garmendia, A. Nogales, and K. Sethi. 2009. "The Burden of Maintenance: Roads in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 14, World Bank, Washington, DC. 310 Africa's Transport Infrastructure Mundy, M., and A. Penfold. 2009. "Beyond the Bottlenecks: Ports in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 13, World Bank, Washington, DC. Roberts, P., K. C. Shyam, and C. Rastogi. 2006. "Rural Access Index: A Key Development Indicator." Transport Paper 10, World Bank, Washington, DC. World Bank. 1998. "China--Forward with One Spirit: A Strategy for the Transport Sector." Research Report 15959, World Bank, Washington, DC. ------. 2010. Railways Database. World Bank, Washington, DC. http:// data.worldbank.org/indicator/IS.RRS.TOTL.KM. CHAPTER 8 Financing: Filling the Gaps The cost of redressing Africa's transport infrastructure needs was esti- mated in chapter 7 as $19.2 billion a year (including the spending by South Africa associated with hosting the World Cup in 2010). Just over half of this amount is for capital investment ($9.64 billion a year), and the rest ($9.55 billion a year) is for operations and mainte- nance (O&M). The overall cost represents 3 percent of Africa's gross domestic product (GDP), with the burden varying greatly by country type. These estimates were based on the connectivity standards of the "base" scenario adopted in chapter 7. While the analysis does involve some bold assumptions, it is believed to be a reasonable basis for esti- mating needs. This chapter presents a detailed analysis of the transport-related rev- enues and expenditures of the sample countries.1 Public spending data are available for only 24 African countries, which collectively account for around 70 percent of African GDP. Therefore, wherever public spending is concerned, totals for country groups and various country types are extrapolated from the available sample based on country GDP. When this analysis is taken together with estimates of investment and maintenance needs from chapter 7, it becomes possible to estimate whether revenues are adequate to meet transport needs. The analysis is 311 312 Africa's Transport Infrastructure done both by individual country and by category, in all cases keeping investment and maintenance separate. Expenditures Expenditures considered include both on-budget and off-budget spend- ing by countries (including state-owned enterprises and extrabud- getary funds), official development assistance (ODA) from the Organisation for Economic Co-operation and Development (OECD) countries, finance from non-OECD countries, and private participation in infrastructure (PPI). Total Spending Total spending on transport infrastructure, including O&M, amounts to about $16.3 billion per year, equivalent to 2.5 percent of GDP on aver- age across all countries. If ODA and OECD finance is excluded, this falls to $13.5 billion, or 2.1 percent of GDP. Middle-income countries account for almost half of total spending, while low-income, fragile states account for less than 5 percent (about $600 million in total). Spending as a percentage of GDP varies substantially by country, rang- ing from less than 1 percent in Chad, to almost 6 percent in Madagascar, to an extreme 13 percent in Cape Verde. Countries such as Chad, Côte d'Ivoire, Nigeria, and Senegal stand out for the very small share of total spending that is allocated to maintenance (appendix 7a contains coun- try details). Country type is an important source of this variation. While transport spending in middle-income states and low-income, nonfragile states is around 3 percent of GDP, resource-rich countries and low-income, fragile states spend only half this proportion of GDP, although for very different reasons (table 8.1). Low-income, fragile states cannot afford to spend significantly on transport infrastructure. For resource-rich countries, transport simply does not seem to be a spending priority. This level of effort by African governments to develop their infrastruc- ture pales in comparison to what the East Asian countries have achieved in recent years. For example, in 2006, China's government invested around 5 percent of GDP in transport infrastructure, not including O&M (Lall, Anand, and Rastogi 2009). This compares with 0.74 percent of GDP directly financed by governments in Africa and 1.34 percent of total GDP invested in transport infrastructure. Table 8.1 Annual Transport Spending by Finance Source and Country Type Percentage of GDP US$ (millions) O&M CAPEX O&M CAPEX Public Public Non-OECD Total Public Public Non-OECD Total Country type sector sector ODA financiers PPI CAPEX Total sector sector ODA financiers PPI CAPEX Total Middle-income 1.88 0.78 0.03 0.01 0.16 0.98 2.86 5,081 2,103 88 22 444 2,657 7,738 Resource-rich 0.32 0.74 0.11 0.34 0.21 1.39 1.72 720 1,646 234 745 469 3,095 3,815 Low-income, nonfragile 0.98 0.67 1.12 0.22 0.12 2.13 3.11 1,084 737 1,241 242 128 2,347 3,431 Low-income, fragile 0.16 0.56 0.61 0.13 0.04 1.33 1.49 60 214 234 9 14 511 571 Total 1.20 0.74 0.28 0.16 0.16 1.34 2.54 7,701 4,724 1,797 1,059 1,055 8,635 16,336 Source: Briceño-Garmendia, Smits, and Foster 2009 (for public spending); PPIAF 2008 (for private flows); Foster and others 2008 (for non-OECD financiers). Note: CAPEX = capital expenditure. Aggregate public sector covers general government and nonfinancial enterprises. Figures are extrapolations based on the 24-country sample covered in phase 1 of the Africa Infrastructure Country Diagnostic (AICD). Totals may not be exact because of rounding errors. 313 314 Africa's Transport Infrastructure Categories of Expenditure Transport spending in Africa is divided almost evenly between invest- ment (1.34 percent of GDP on average over all countries) and O&M (1.20 percent of GDP on average). But the composition of spending varies substantially among country groups. Middle-income countries allocate two-thirds of transport spending to maintenance, which proba- bly reflects the fact that they have already built much of the infrastruc- ture they need. By contrast, all other country groups allocate at least two-thirds of their transport spending to capital investment. Sources of Funding The public sector is by far the most important source of finance for trans- port spending in Africa. Overall, about three-quarters of transport infra- structure spending is provided by domestic public institutions. In the middle-income countries, domestic public sector resources (comprising tax revenues and user charges raised by state entities) account for almost all finance for transport spending. In resource-rich countries, about 60 percent comes from domestic public institutions, but this falls to about half in the low-income countries. In the middle-income states and low-income, nonfragile states, domestic public spending is focused on maintenance, whereas in the resource-rich states and low-income, fragile states, it is focused on capital spending. External finance is primarily for investment--including asset rehabilitation and reconstruction--and in most cases does not provide for O&M. Overall, transport (roads in particular) is the most expensive type of infrastructure in general government accounts. Expenditures range from about half of all general government spending on infrastructure in middle- income countries to 80 percent in low-income, fragile countries. From their central government budgets alone, African countries allocate an average of 0.7 percent of GDP to spending on transport infrastructure (table 8.2). As a percentage of GDP, budget spending on transport infrastructure is com- parable across low- and middle-income countries. Given the much higher GDP of the middle-income countries, however, absolute spending per capita in these countries can be several times higher than in the low-income countries. Table 8.2 looks only at on-budget transport spending (the sum of on-budget operating expenditure [OPEX] and capital expenditure [CAPEX]). In other words, these are annual budgetary flows. In contrast, table 8.1 looks at total public spending (on-budget and off-budget). For capital investment expenditures in transport infrastructure, the domestic public sector also dominates in both middle-income and Financing: Filling the Gaps 315 Table 8.2 Transport Spending as a Share of Total Budget in Africa Percentage of GDP US$ (millions) Total Total Country type Transport budget Transport budget Middle-income 0.69 1.59 1,872 4,297 Resource-rich 0.83 1.67 1,836 3,711 Low-income, nonfragile 0.73 1.50 803 1,655 Low-income, fragile 0.58 0.71 224 271 Total 0.73 1.56 4,691 10,010 Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­6. Averages weighted by country GDP. Figures are extrapolations based on the 24-country sample covered in AICD phase 1. Totals may not be exact because of rounding errors. resource-rich countries. But in total, external finance contributes roughly half of Africa's total capital spending on transport infrastructure. External sources include ODA from the OECD countries, official finance from non-OECD countries (such as China, India, and the Arab states), and PPI. ODA is by far the largest source of external finance for transport infra- structure in Africa, accounting for half the total. Non-OECD finance and PPI each account for a quarter of external finance (figure 8.1). Nearly one-quarter of the capital investment in resource-rich countries is financed from non-OECD sources (mostly China). ODA plays a substan- tial role everywhere except in the middle-income countries. Only a hand- ful of countries (in particular, Mozambique, Nigeria, and Uganda) enjoy a significant contribution from the private sector. Although South Africa is by far the largest recipient of private transport investment in absolute terms, this total is only a small percentage of the country's public spend- ing on transport infrastructure. The sources of external finance exhibit clear patterns of specializa- tion. For example, ODA is concentrated on road infrastructure. Much non-OECD finance has gone to the rail sector, due to the synergies with natural resource sectors that are an important focus of coopera- tion among developing countries. PPI has benefited only a small seg- ment of high-volume toll roads but has contributed significantly to railways and ports. ODA for capital investment in transport is particularly significant in the low-income states, with the nonfragile states receiving higher levels of support relative to their GDP (around 1.1 percent) than the fragile states (only 0.6 percent). PPI for transport has tended to go to middle-income and resource-rich countries, which have the greatest 316 Africa's Transport Infrastructure Figure 8.1 Sources of Finance for Transport Spending in Africa by Country Madagascar Mozambique Congo, Rep. Zambia Ethiopia Malawi Lesotho Mali Tanzania Kenya South Africa country Ghana Benin Botswana Niger Senegal Uganda Nigeria Namibia Cameroon Rwanda Congo, Dem. Rep. Côte d'Ivoire Chad 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 percentage of GDP public ODA non-OECD financiers PPI Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­06. Because numbers for Cape Verde are exceptionally high, at around 13 percent of GDP, the country is excluded from this figure to better represent the situation in the other countries. ability to pay for services. Both groups have received around 0.2 percent of GDP annually in private capital for transport in recent years. Non- OECD finance for transport is more important than PPI everywhere but in the middle-income countries. This support is highest for the resource- rich countries (0.34 percent of GDP) but remains significant even for the fragile states (in excess of 0.10 percent of GDP) (figure 8.2). The Balance between Investment and Maintenance The balance of investment and maintenance expenditure varies signifi- cantly among countries (figure 8.3). Discounting the Democratic Republic Financing: Filling the Gaps 317 Figure 8.2 Sources of Funding for Transport Infrastructure Capital Investment 2.5 percentage of GDP 2.0 1.5 1.0 0.5 0 IC le a h le ric ric gi gi M Af fra ra - ce nf ur C, no LI so C, re LI region PPI non-OECD ODA public Source: Briceño-Garmendia, Smits, and Foster 2009 (for public spending and ODA); PPIAF 2008 (for private flows); Foster and others 2008 (for non-OECD financiers). Note: MIC = middle-income country; LIC = low-income country. of Congo, which spends little on either, and with the exception of Cameroon and Kenya, only the middle-income countries (South Africa and Namibia) spend more on maintenance than on investment. The Balance between Finance Needs and Commitments Sources of finance are not completely fungible (for instance, ODA funds for road rehabilitation may not be available for routine maintenance). Therefore, spending on maintenance (figure 8.4, panel a) and investment expenditures (figure 8.4, panel b) are examined separately. For purposes of comparison, the graphics are standardized based on the needs of each coun- try as estimated in chapter 7. The sources of revenue to meet these needs are presented as cumulative bars in the graphs, grouped by country type. A quarter of the countries (Benin, Botswana, Cameroon, the Republic of Congo, Kenya, Lesotho, and Malawi) appear to provide adequately for maintenance. Countries with good maintenance provisions are found in each of the four country types. Some of these look anomalous. For exam- ple, Kenya's maintenance expenditure appears to be adequate, and allow- ing for the inclusion of ODA, investment expenditure appears sufficient to meet assessed needs. Yet only 50 percent of its network is in good condition, and 34 percent is in poor condition (see appendix 2i). Similarly, Cameroon appears to be spending enough on maintenance (though not on investment), despite the fact that only 36 percent of its network is in good condition. These findings highlight the fact that coun- tries may be currently committing enough to maintenance (which is good 318 Africa's Transport Infrastructure Figure 8.3 Capital Investment and O&M Spending for Transport from All Sources by Country Madagascar Mozambique Congo, Rep. Zambia Ethiopia Malawi Lesotho Mali Tanzania Kenya South Africa country Ghana Benin Botswana Niger Senegal Uganda Nigeria Namibia Cameroon Rwanda Côte d'Ivoire Chad Congo, Dem. Rep. 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 percentage of GDP O&M capital Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­06. Because numbers for Cape Verde are exceptionally high, at around 13 percent of GDP, the country is excluded from this figure, offering a better representation of the situa- tion in the other countries. news), but have road networks not in good condition because of large backlogs still to be overcome. For many countries, however, the shortfalls are very significant. The worst overall deficiencies are found in the low-income, fragile countries. While a number of countries of each country type appear to be spending enough overall to meet the estimated capital investment requirements, Botswana, Cape Verde, the Republic of Congo, Nigeria, Senegal, and Zambia are the only countries able to cover their estimated investment needs without ODA. Even in these countries, investment expenditure Financing: Filling the Gaps 319 Figure 8.4 Transport Spending as Percentage of Needs a. Maintenance spending as percentage of maintenance needs Botswana Lesotho MIC South Africa Namibia Malawi Kenya Benin Ghana Madagascar LIC, nonfragile Tanzania Uganda Rwanda country Mozambique Ethiopia Senegal Niger Mali Cameroon resource-rich Congo, Rep. Zambia Nigeria Chad LIC, fragile Congo, Dem. Rep Côte d'Ivoire 0 50 100 150 200 250 300 percentage of needs (continued) 320 Africa's Transport Infrastructure Figure 8.4 (continued) b. Investment spending as percentage of investment needs Botswana Lesotho MIC South Africa Namibia Ethiopia Senegal Rwanda Malawi Kenya LIC, nonfragile Mali Ghana country Mozambique Benin Madagascar Uganda Niger Tanzania Nigeria LIC, fragile resource-rich Congo, Rep. Zambia Cameroon Chad Côte d'Ivoire 0 50 100 150 200 250 300 percentage of needs public investment non-OECD PPI ODA Source: Briceño-Garmendia, Smits, and Foster 2009. Note: LIC = low-income country; MIC = middle-income country. Numbers for Cape Verde are exceptionally high, surpassing 600 percent in both cases. Accordingly, it is excluded from this figure and from figure 8.5 to offer a better representation of the situation in the other countries. The Democratic Republic of Congo is omitted because of the incompleteness of the public finance data for this country. may not be excessive for two reasons. First, as explained in chapter 7, some types of investment needs (bridges, for example) were not included in the estimate, and there is a range of variability in unit costs not taken into account in the calculations. Second, these countries may simply be choosing to eliminate rehabilitation backlogs more quickly than assumed Financing: Filling the Gaps 321 in the model. Overall, the inability of many countries to finance both maintenance and capital requirements on their own, or with commercial borrowings, has been disguised by heavy dependence on ODA for capi- tal investment. For a better assessment of the burden of these shortfalls, the same basic data can be shown as percentages of GDP (figure 8.5a and figure 8.5b). Figure 8.5 Needs and Spending as Percentage of GDP a. Maintenance Botswana Lesotho South Africa MIC Namibia Malawi Kenya LIC, nonfragile Benin Ghana Madagascar Tanzania Uganda Rwanda country Mozambique Ethiopia Senegal Niger Mali Cameroon resource-rich Congo, Rep. Zambia Nigeria Chad LIC, fragile Congo, Dem. Rep. Côte d'lvoire 0 1 2 3 4 5 6 7 percentage of GDP needs spending (continued) 322 Africa's Transport Infrastructure Figure 8.5 (continued) b. Investment Botswana Lesotho South Africa MIC Namibia Ethiopia Senegal Rwanda Malawi LIC, nonfragile Kenya Mali Ghana country Mozambique Benin Madagascar Uganda Niger Tanzania Nigeria resource-rich Congo, Rep. Zambia Cameroon Chad LIC, fragile Côte d'lvoire 0 1 2 3 4 5 percentage of GDP needs spending spending w/o ODA Source: Briceño-Garmendia, Smits, and Foster 2009. Note: MIC = middle-income country; LIC = low-income country. Numbers for Cape Verde are exceptionally high; the country is excluded from this figure to better represent the situation in the other countries. The Democratic Republic of Congo is omitted because of the incompleteness of the public finance data for this country. What Can Be Done about the Shortfalls? Closing Africa's transport infrastructure funding gap will inevitably require both reforms to reduce inefficiencies and the creation of a more attractive investment climate for external finance. Financing: Filling the Gaps 323 Efficiency Improvements Inefficiencies in transport spending are estimated to cost countries in Africa a total of about $4.1 billion a year. Addressing these ineffi- ciencies would make an additional $2.4 billion of resources per year directly available to the sector institutions (as explained later in this chapter), as well as benefiting road users substantially. In some countries, these additional resources would completely eliminate spending shortfalls. In other countries (particularly the low-income, fragile states), however, a sizeable funding gap would remain. For the low-income, fragile states as a group, the funding gap would remain greater than $2.4 billion a year even if all spending inefficiencies were eliminated. Three main opportunities for eliminating inefficiencies have been identified. First, raising user charges closer to maintenance cost-recovery levels would provide more efficient price signals and increase road agency revenues. Second, improving budget execution rates would more fully exploit resources allocated to public investment. Third, a higher allocation of resources to asset maintenance would substan- tially improve efficiency both by preventing costly rehabilitation (par- ticularly for roads) and by yielding direct benefits to the users of infrastructure. In the first two cases, the efficiency gains directly benefit the government. The third reform would primarily benefit infrastructure users, although the government may also benefit (see appendix 7b for country-level calculations). In addition, there are other possibilities of reducing expenditures through reallocating sector funding, lowering the standards of provision, and reducing backlogs in maintenance spending. Improved Collection of User Charges In the road sector, lending institutions and governments are moving to employ indirect user charges such as fuel levies and taxes to cover road maintenance costs (see chapter 2). On that principle, fuel levies must be sufficiently high to cover the full maintenance costs imposed by the use of the road network. Currently, even if fuel levies were fully col- lected, only one-third of African countries could cover their mainte- nance costs. Undercollection of fuel levies for road sector maintenance is also an inefficiency, albeit a relatively small one. Underpricing for road use--that is, setting fuel levies too low to cover road network maintenance--is the more pressing issue, resulting in lost revenues estimated at $1.4 billion a year. 324 Africa's Transport Infrastructure Improved Capital Fund Disbursement Central governments are key players in infrastructure investment. Inefficiencies within the public expenditure management systems are therefore particularly significant. African countries, on average, fail to spend as much as one-fifth of their capital budgets for transport (table 8.3). The poor timing of project appraisals and late releases of budgeted funds because of procurement problems often prevent the use of resources within the budget cycle. Delays affecting fund releases within the budget year are also associated with poor project preparation, leading to changes in the initial terms agreed upon with contractors (such as deadlines, tech- nical specifications, budgets, and costs). In other cases, funds are reallo- cated from transport to nondiscretionary spending driven by political or social pressures. Historically, the road sector has had the greatest difficulty spending the amounts allocated to it--sometimes as much as 60 percent of the budget goes unspent. Only about 80 percent of the capital budget allocation for transport infrastructure is actually used. About $1.3 billion in public investment ear- marked for the transport sector is therefore diverted elsewhere. If these inefficiencies were eliminated, countries could increase their capital spend- ing on transport by an average of 25 percent without any increase in budget allocations. This assertion assumes that funds will reach their intended destinations, which is not always the case (Reinikka and Svensson 2002). It also assumes that budget estimates are realistic and aligned with resource availability. Nevertheless, the planning, budgeting, and procurement chal- lenges associated with unused allocations should be central to the region's reform agenda. Improving disbursement will not be easy. A principal cause of under- execution of capital budgets is overoptimistic budgeting resulting from inadequate absorptive capacity and weak sector policy. Improving budget Table 8.3 Average Percentage of Capital Budget Actually Spent, by Country Type All Transport Country type infrastructure infrastructure Middle-income 78 100 Resource-rich 65 73 Low-income, nonfragile 76 72 Low-income, fragile -- -- Africa 75 79 Source: Adapted from Briceño-Garmendia, Smits, and Foster 2009. Note: -- = not available. Based on annualized averages for 2001­06. Financing: Filling the Gaps 325 planning and expenditure forecasting in the road sector will thus require more discussion between the relevant transport ministry, min- istries of finance and planning, and donors in the design of a national transport policy. Better Allocation of Funds to Maintenance On average, 30 percent of African infrastructure assets are in need of rehabilitation because of neglected maintenance in the past (figure 8.6). The cost of operating a vehicle on an uncongested road depends on the type of vehicle, the physical layout of the road, and the condition of the road. All other factors constant, the cost per vehicle-kilometer increases as the condition of the road deteriorates, with the rate of cost increase dependent on the vehicle and the road. The rate of road deterioration depends on traffic volumes (particularly the volume of heavy vehicles), topography, and climate. For paved roads, routine maintenance--including filling of potholes and patching--can slow deterioration. According to recent evidence con- cerning costs in the region, such maintenance should cost no more than $2,000 per kilometer per year for a two-lane road. Periodic maintenance in the form of surface overlays can return the road surface to its original condition if undertaken in a timely manner. This will probably cost about $20,000 per kilometer per treatment, and is likely to be required every five or six years. If routine and periodic maintenance are not performed, road deterioration will continue to the point of structural failure, which Figure 8.6 Transport Infrastructure Assets in Need of Rehabilitation 50 average rehabilitation 40 index (%) 30 20 10 0 s er ure s s ad ay ad e ilw av uct ro ro ag n al ra r st ai r ru fra m in Source: Carruthers, Krishnamani, and Murray 2009. 326 Africa's Transport Infrastructure can only be rectified by complete rehabilitation. This will probably cost at least $250,000 per kilometer. Timely routine and periodic maintenance defers--potentially indefinitely--the need for complete rehabilitation. Deterioration of gravel and earth roads has equivalent effects on vehicle operating costs, although they require different treatment strategies with distinct cost profiles. A wide range of treatments is possible, and the opti- mal strategy may vary substantially from case to case. Underspending on routine and periodic maintenance is a major waste of resources for two reasons. First, there may be a cost to the road agency because the present value of completely rehabilitating roads is greater than the present value of sound preventive maintenance. The potential efficiency saving from adequate maintenance depends on the type of road, traffic volumes, and the discount rate. If the discount rate (that is, the opportunity cost of capital) is high, improved maintenance may not provide significant cost savings to the road agency. Second, road users, however, certainly bear the higher operating costs of roads in a poor state of repair, which usually cancels any benefit to the road agency. Using the Highway Development and Management Model (HDM4), the costs of two maintenance and rehabilitation alternatives for road agencies were compared across several typical road classes in Africa. The alternatives were to (i) perform annual routine maintenance and rehabilitate the road when its condition became poor, and (ii) perform annual routine maintenance and proper periodic maintenance. Sixteen road classes were evaluated, based on two climate/terrain types, four traf- fic levels, and two road condition types. (The details of the analysis are presented more fully in appendix 7c.) According to the analysis, performing timely periodic maintenance is likely to reduce agency costs if no discounting is applied, except where traf- fic volumes are very low (under 500 vehicles per day). At a 12 percent dis- count rate, whether timely routine maintenance reduces total agency costs depends on a combination of climate/terrain, traffic volumes, and initial surface condition. But when user costs are taken into account, the only sit- uation in which timely periodic maintenance produces a net loss is when traffic volume is low and the initial condition is good. As surface condition deteriorates from good to fair, however, timely routine maintenance will once again produce a net benefit. Moreover, only a small proportion of road traffic meets the criteria that result in a net loss. While this analysis consid- ers only a small set of the wide range of possible road scenarios, it can be safely concluded that timely periodic maintenance will yield high net social benefits regardless of whether the road agency benefits. Financing: Filling the Gaps 327 Accurately determining the respective benefit to the road agency and road users of more efficient maintenance would require a more detailed comparison of current and optimal maintenance strategies in each country. But as the benefit to road users outweighs the benefit to the road agency in most cases, the analysis that follows emphasizes the total benefit to society rather than just to the road agency. The approach adopted uses the Road Network Evaluation Tool (RONET) analyses discussed in chapter 2 to compare the road-user cost savings achieved by better maintenance to the agency cost of the chosen maintenance strategy (figure 8.7). The basic calculation assumes that countries attempt to secure the cus- tom standard of maintenance. This involves applying a high standard of maintenance to the primary roads, a medium standard to the secondary roads, and a low standard to the tertiary roads (as discussed in chapter 2). This approximates the pragmatic scenario standards for roads discussed in chapter 7. The net benefit-cost ratios of applying a high standard of main- tenance to the entire road network are very similar to those of the custom standard. This is because improving maintenance beyond the custom stan- dard has further benefits on some links; meanwhile, on other routes with low traffic volume, the costs of a high standard of maintenance exceed the benefits of reduced vehicle-operating costs. Only where the maintenance programs are optimized (that is, where the optimal treatment is deter- mined for each road on a link-by-link basis) is the ratio substantially higher. The custom standard calculation may therefore be viewed as a lower bound and the optimal standard calculation as an upper bound for the potential efficiency gains of an improved maintenance strategy. The ratio of user cost savings to agency expenditures range between 1.1:1.0 (Nigeria) and 5.4:1.0 (Tanzania) for the custom standard and between 3.5:1.0 (Madagascar) and 8.8:1.0 (Cameroon) for the optimal standard (figure 8.7). The total agency and vehicle-operating cost savings per country for each scenario are estimated by multiplying the cost sav- ings ratio by the estimated magnitude of current maintenance under- spending (table 8.4). Given these ratios, priority must be given to maintenance expenditures whether that policy is immediately beneficial to the road agency or not. Other Reallocations in the Transport Sector Comparing actual expenditures with estimated needs suggests that some countries spend more than the amount that would ostensibly cover their needs. The sum of overspending in these countries is about Figure 8.7 Efficiency Gains from Improved Maintenance 328 10 present value impact of increments in maintenance spending 9 8.8 8.6 8 7.3 7.1 7 6.1 6.3 6.0 6 5.6 5.4 5.0 5 4.6 4.8 4.8 4.5 4.5 4.5 4.3 4.4 4.5 4.2 4.3 4.1 4.2 3.9 3.9 4.0 3.8 3.9 4 3.5 3 2.8 2.7 2.6 2.4 2.7 2.6 2.5 2.7 2.5 2.6 2.2 2.1 2.2 2.1 2.2 2.1 2.3 2.4 2.2 2.1 2.1 2 1.9 1.9 1.9 1.7 1.6 1.6 1.3 1.3 1.4 1.2 1.3 1.1 1.3 1 0 n o n ad re ia a a ad ho r i Na e ia r ria da l a ia da a aw ga ca ge an ny ric bi qu as ni oo op ib an oi ge Ch an an t as ne m Be Ni aF so al Af Ke Gh m Iv bi er nz hi Rw Ug Ni Za ag M d' Se Le am m in h Et Ta ut te Ca rk oz So Cô Bu M M very high standard optimal standard custom standard Source: Calculation by A. Nogales (based on data from Gwilliam and others 2009). Note: The graph shows the present value impact in terms of the ratio of savings of user costs to unit increases of maintenance expenditures across three maintenance standards. Financing: Filling the Gaps 329 Table 8.4 Overall Value of Agency and Vehicle-Operating Cost Savings (US$ millions) Agency and Agency and user Potential Potential user savings per dollar savings per dollar savings, savings, spent on maintenance, spent on maintenance, custom optimum Country custom standard optimum standard standard standard Benin 3.3 6.1 0 0 Botswana 1.0 5.0 0 0 Cameroon 3.5 8.8 89 220 Chad 1.6 6.3 36 139 Côte d'Ivoire 1.2 4.1 50 172 Ethiopia 1.1 8.6 68 525 Ghana 1.7 4.2 119 315 Kenya 1.4 4.6 0 0 Lesotho 3.5 7.3 17 37 Madagascar 0.3 3.5 4 49 Malawi 0.9 3.8 7 27 Mozambique 0.9 3.9 37 160 Namibia 1.6 4.5 25 72 Niger 0.1 3.9 3 98 Nigeria 1.7 4.2 459 1,134 Rwanda 1.1 5.6 1 6 Senegal 0.3 4.5 4 59 South Africa 0.6 4.3 164 1,173 Tanzania 4.4 7.1 146 234 Uganda 1.6 6.0 111 420 Zambia 0.4 3.9 0 0 Source: Calculation by A. Nogales (based on data from Gwilliam and others 2009). $1.9 billion a year. This spending--funded through public budgets-- includes domestically raised funds and international aid (OECD and non- OECD sources). Most of it is driven by apparent overinvestment in some road networks, even as other roads in the same countries are undermain- tained. In nine countries, the potential for reallocating this excess spending looks particularly substantial--in excess of 0.5 percent of GDP (figure 8.8). The most dramatic example is Cape Verde, whose excess spending is equivalent to more than 10 percent of GDP; Ethiopia, the Republic of Congo, Botswana, and Lesotho all have excesses around 1 percent of GDP. It is possible that such overspending reflects a political decision to rapidly address a rehabilitation and investment backlog in the road sector, compressing a large spending program into a shorter time than is assumed in the model on which the estimates are based (see chapter 7). Neverthe- less, in some cases this money might be better spent elsewhere. 330 Africa's Transport Infrastructure Figure 8.8 Potential for Reallocation of Spending Cape Verde Ethiopia Congo, Rep. Botswana Lesotho Zambia Nigeria country Malawi Senegal Rwanda Kenya Cameroon Benin Mali Ghana Mozambique 0 0 5 0 5 0 5 0 5 0 5 0 5 0 5 0 5 .0 11 5 11 0 12 5 .0 0 5 5 1. 2. 2. 3. 3. 4. 4. 5. 5. 6. 6. 7. 7. 8. 8. 9. 9. . . . 0. 1. 10 10 percentage of GDP Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­06. Estimates of the economic rates of return to key infrastructure projects are helpful in evaluating whether the apparent overspending on invest- ment is beneficial. Across infrastructure interventions in Africa, the eco- nomic rates of return to road maintenance are the highest, averaging more than 100 percent--well above the typical rates of return for road rehabil- itation and upgrading (table 8.5). Hence, countries that overspend on capital investment while underspending on maintenance do not appear to be allocating resources efficiently. Lower Standards To reach a given infrastructure connectivity target (as defined in chapter 7), a wide array of technical alternatives may be pursued, each with a distinct cost and quality of service. Where budgets are constrained, policy makers must choose between providing a high level of service to a relatively small cross-section of the population or a lower level of service to a larger cross- section. Thus, providing a high level of service may not be in a country's best interest. Financing: Filling the Gaps 331 Table 8.5 Economic Rates of Return for Key Infrastructure Interventions in Africa Railway Road Road Road Power Country type rehabilitation rehabilitation upgrades maintenance Irrigation generation Water Middle-income 18.5 45.4 19.8 143.0 19.3 13.6 26.8 Resource-rich 10.8 16.2 17.4 114.5 24.2 20.2 37.0 Low-income, nonfragile 6.2 17.6 12.8 125.7 17.2 14.3 7.7 Low-income, fragile 2.5 9.2 12.0 67.6 -- 24.7 36.9 Total 5.1 24.2 17.0 138.8 22.2 18.9 23.3 Source: Derived from Foster and Briceño-Garmendia 2009. Note: -- = not available. The availability of cost-saving technologies varies considerably among transport sectors. In the case of roads, for example, the costs of reaching road connectivity targets vary depending on the chosen engineering stan- dards. For the base scenario considered in this analysis, regional and national connectivity is achieved by a good-condition asphalt road net- work, with at least two lanes for regional connectivity and at least one lane for national connectivity. If a single-surface-treatment road in fair condition were substituted for an asphalt road in good condition, the cost of reaching the same connectivity targets could be reduced by 30 percent. Similarly, relaxing the standards for rural road construction allows a greater level of connectivity to be achieved with a given budget. Addressing Rehabilitation Backlogs The investment needs presented in this book are based on the objective of redressing Africa's infrastructure backlog within 10 years. Middle- income, resource-rich states and low-income, nonfragile states could meet this target within existing resource envelopes (including existing levels of ODA) if inefficiency was substantially reduced. The same cannot be said for the low-income, fragile states. For a few of these states, the infrastruc- ture backlog is so great that policy makers should consider taking more time to attain targets, using lower-cost technologies, or both. Even the low-income, fragile states could attain their transport infrastructure tar- gets without increasing their spending envelopes if they were willing to take 40 years instead of 10 to address their investment backlogs (assum- ing full elimination of inefficiencies). Extending the time horizon for the achievement of these goals should make the targets more affordable. But how long would the extension need to be to make the infrastructure targets attainable without 332 Africa's Transport Infrastructure increasing existing spending envelopes? Figure 8.9 shows that by spreading the investment needs over 45 years rather than 10, low-income, fragile states could achieve their proposed targets within existing spending envelopes. Without efficiency gains, however, these countries would require much more than 45 years to meet their infrastructure targets or, alternatively, would need to double their spending to reach their target in 45 years. Advocating that countries delay in making up backlogs, however, is not the same as advocating that they defer current and periodic maintenance. The latter would be a false economy, which would eventually increase the size and cost of the backlog. Figure 8.9 Cost of Redressing Infrastructure Backlogs over Varying Time Frames a. Resource envelope plus potential efficiency gains variation in resources needed 600 (% deviation from current 500 400 envelope) 300 200 100 0 20 24 26 28 30 32 34 36 38 40 42 44 48 10 12 14 16 18 22 46 50 number of years needed to attain funding targets b. Existing resource envelope variation in resources needed 600 (% deviation from current 500 400 envelope) 300 200 100 0 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 number of years needed to attain funding targets LIC fragile other Source: Africa Infrastructure Country Diagnostic. Note: LIC = low-income country. LIC nonfragile, MICs (middle-income countries), and resource-rich countries show almost identical patterns and, therefore, are taken as the single category "other." Financing: Filling the Gaps 333 In sum, governments could directly capture $2.4 billion by improving transport infrastructure management and institutions (table 8.6). Raising fuel levies to cost-recovery levels would secure $600 million per year. Potential savings from improved levy collection amount to $500 million, and $1.3 billion per year could be saved by raising capital budget execution rates through improvements to the preparation and procurement of projects. The greatest potential gain would come from reductions in vehicle-operating costs resulting from better road conditions. Road users would be the primary beneficiaries, saving an estimated $1.34 billion to $4.84 billion. These savings, however, would necessitate an increase in agency expenditures; therefore, they must be reduced by the additional maintenance expenditure required to realize them before being added to agency savings. Looking across countries (see appendix 7d), about 10 countries could realize efficiency savings of 0.5 percent of GDP or more--and more than 1.0 percent in the case of Namibia and Malawi (figure 8.10). Nigeria has Table 8.6 Potential Gains from Improved Efficiency in Transport Spending (US$ millions annually) Low-income, Low-income, Middle-income Resource-rich nonfragile fragile countries countries countries countries Total Infrastructure spending needs 8,430 3,810 3,797 3,155 19,193 Spending directed to needs 7,738 3,113 3,266 571 16,336 Gain from eliminat- ing inefficiencies 688 788 541 107 2,368 - Gain from raising capital budget execution 613 455 162 61 1,298 - Gain from cost recovery through fuel levy 38 163 211 46 574 - Gain from undercollection 36 170 169 -- 497 (Financing gap) or surplus (4) 91 10 (2,477) (489) Source: Briceño-Garmendia, Smits, and Foster 2009. Note: -- = not available. Totals may not be exact because of rounding errors. 334 Africa's Transport Infrastructure Figure 8.10 Potential Efficiency Gains from Different Sources Namibia Malawi Ethiopia Mozambique Niger Lesotho Zambia Madagascar Ghana Congo, Rep. Tanzania country Chad Côte d'lvoire Cape Verde Botswana Cameroon Senegal Kenya South Africa Nigeria Rwanda Uganda Benin 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 percentage of GDP undercollection capital execution tariff cost recovery Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­06. potential savings of around 0.2 percent of GDP, amounting to about $200 million a year. For most countries, insufficient expenditure on road maintenance and excessive user costs are the main sources of inefficiency. But in a few countries (such as Zambia, Ethiopia, and the Republic of Congo), low capital-budget execution is the most pressing issue. Increasing Funding Even if they can improve the efficiency with which current funding lev- els are employed, many countries will also require additional sources of Financing: Filling the Gaps 335 funds to meet their spending needs. But prospects for increased funding for transport from all sources are limited (World Bank 2009). Domestic public finance is the largest source of funding today, but it presents little scope for an increase, except possibly in countries enjoying natural resource windfalls. ODA to African transport infrastructure has grown substantially in recent years, in line with political pledges, but this assis- tance could slow down because of fiscal pressures in donor countries related to the 2008 economic downturn. Non-OECD finance has also been rising steeply, but its future is now unclear. Private participation, also very buoyant during Africa's recent growth upswing, will be particularly vulnerable to the downturn in global markets. Finally, local capital mar- kets have so far contributed little to infrastructure finance outside South Africa, though they could eventually become more important in some of the region's larger economies. These last two sources of funding are of lim- ited relevance to the road sector in low-income, fragile states, where fund- ing gaps are most significant, because of road users' limited ability to pay. Raising More Domestic Public Finance A key question is the extent to which countries may be willing to allocate additional fiscal resources to infrastructure. Prior to the current financial crisis, the fiscal situation in Africa was favorable. Rapid economic growth--averaging 4 percent per year from 2001 to 2003 and 5 percent a year from 2004 to 2008--translated into increased annual domestic fis- cal revenues of just over 3 percent of GDP on average. In resource-rich countries, burgeoning resource royalties added 7.7 percent of GDP to the public budget. In the low-income countries, substantial debt relief increased external grants by almost 2 percent of GDP. Nevertheless, a surprisingly small portion of the additional resources was allocated to infrastructure (table 8.7). This was especially true in the resource-rich countries, particularly Nigeria. Huge debt repayments absorbed the fiscal windfalls in these countries. As a result, budgetary spending contracted by 3.7 percent of GDP, with infrastructure invest- ment falling by almost 1.5 percent of GDP. In the middle-income coun- tries, budgetary spending increased by almost 4.1 percent of GDP, but the additional resources went primarily to social sector spending, and the effect on infrastructure spending was almost negligible. Only in the low- income countries did the overall increases in budgetary expenditure have some effect on infrastructure spending. Even there, however, the effect was fairly modest and confined to capital investment. The low-income, nonfragile countries allocated 30 percent of their budgetary increase to 336 Africa's Transport Infrastructure Table 8.7 Net Change in Central Government Budgets by Economic Use, 1995­2004 Percentage of GDP Low-income, Low-income, Middle-income Resource-rich nonfragile fragile Use Africa countries countries countries countries Increase (decrease) in net expenditure budget 1.89 4.08 (3.73) 1.69 3.85 Increase in current infrastructure spending as a share of expenditures 0.00 0.02 0.03 0.00 0.09 Increase (decrease) in capital infrastructure spending as a share of expenditures (0.14) 0.04 (1.46) 0.54 0.22 Source: Adapted from Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­06. Averages weighted by country GDP. Totals are extrapolations based on the 24-country sample covered in AICD phase 1. infrastructure investments. The fragile states, despite seeing their overall budgetary expenditures rise by about 3.9 percent of GDP, devoted only 6 percent of the gain to infrastructure. Compared with other developing regions, Africa does a particularly poor job of collecting tax revenues, thus limiting its public financing capa- bilities. Domestic revenue generation is around 23 percent of GDP, trail- ing averages for other developing countries. That figure is lowest for the low-income countries (less than 15 percent of GDP per year). Despite strong growth in the past decade, domestically raised revenues increased by less than 1.2 percent of GDP. This suggests that challenging institu- tional reforms will be required to increase the effectiveness of revenue collection and broaden the tax base. The capacity of African countries to borrow from domestic and exter- nal sources is also limited. Domestic borrowing is often very expensive, with interest rates far exceeding those on concessional external loans. Because of the scarcity of private domestic savings, public domestic bor- rowing tends to precipitate sharp increases in interest rates, particularly for the poorest countries. For many African countries, the ratio of debt service to GDP is more than 6 percent. The ongoing financial crisis is expected to reduce fiscal receipts glob- ally, and Africa will not be exempt. Growth projections for the coming Financing: Filling the Gaps 337 years have been revised downward from 5.1 percent to 3.5 percent, which will reduce tax revenues and likely depress demand and willing- ness to pay for infrastructure services. Commodity prices have fallen to levels of the early 2000s. The effect on royalty revenues, however, will depend on the savings policies of each country. A number of oil produc- ers have made a practice of saving royalty revenues collected from sales at prices above $60 a barrel, so the downturn will affect their savings accounts more than their budgets. In addition, many African countries are devaluing their currency, reducing the purchasing power of domestic resources. Overall, the global financial crisis will put substantial pressure on public sector budgets. A further perverse influence should be noted. Based on recent global experience, fiscal adjustment episodes tend to disproportionately affect public investment--infrastructure in particular. During earlier crises in East Asia and Latin America, infrastructure spending--especially road maintenance--was vulnerable to budget cutbacks. Cuts in infrastruc- ture investment in eight Latin American countries amounted to an average of 40 percent of the observed fiscal adjustment between the early 1980s and the late 1990s (Calderón and Servén 2004). This reduc- tion was remarkable because public infrastructure investment already represented less than 25 percent of overall public expenditure in Latin American countries. These infrastructure investment cuts were later identified as the underlying problem holding back economic growth throughout the region during the 2000s. Similar patterns were observed in East Asia during the financial crisis of the mid-1990s. For example, Indonesia's total public investment in infrastructure dropped from between 6 and 7 percent of GDP in 1995­97 to 2 percent in 2000. Given recent spending patterns, changes in the overall budget envelope will likely affect infrastructure investment in Africa in a similar pro- cyclical manner. Official Development Assistance: Sustaining the Scale-Up For most of the 1990s and early 2000s, ODA for transport infrastructure in Africa remained steady at around $1.3 billion a year. The launch of the Commission for Africa Report in 2004 was followed in July 2005 by the Group of Eight Gleneagles Summit, where the Infrastructure Consor- tium for Africa was created to focus on scaling up donor finance to meet Africa's infrastructure needs. Donors have so far kept their promises: ODA flows to African transport infrastructure increased by more than 50 percent between 2004 and 2007--from $2.0 billion to $3.2 billion. 338 Africa's Transport Infrastructure More than 80 percent of this ODA comes from multilateral donors--the African Development Bank (AfDB), the European Commission, and the International Development Association (IDA) of the World Bank Group. Among bilaterals, France and Japan have made significant contributions. A significant lag occurs between ODA commitments and their disburse- ment, suggesting that disbursements should continue to increase in the coming years. The commitments reported above are significantly higher than the estimated ODA disbursements of $1.8 billion (see table 8.1). This gap reflects delays typically associated with project implementation. Because ODA is channeled through the government budget, the execu- tion of funds faces some of the same problems affecting domestically financed public investment, including procurement delays and low administrative capacity. Differences between the financial systems of the donor country and the receiving country, as well as unpredictability in the release of funds, may further delay the disbursement of resources. Nevertheless, if all commitments up to 2007 are fully honored, ODA dis- bursements should rise significantly (IMF 2009). ODA was set to increase further before the global economic crisis, but prospects are no longer as bright. The three multilateral agencies--the AfDB, the European Commission, and the IDA--secured record replen- ishments for their concessional funding windows for the three to four years beginning in 2008. In principle, the multilateral agencies could pro- vide $5.2 billion per year for African infrastructure in the near future, with transport receiving a substantial share of that amount. In practice, however, the crisis may divert multilateral resources from infrastructure projects to emergency fiscal support. Historical trends suggest that ODA has tended to be procyclical rather than countercyclical (IMF 2009; ODI 2009). Bilateral support, based on annual budget determinations, may be more sensitive to the fiscal squeeze in OECD countries, and some decline can be anticipated. Non-OECD Financing: Will Growth Continue? Non-OECD countries financed about $1.1 billion of African transport infrastructure annually between 2001 and 2006 (see table 8.1). This is substantially less than what was provided by ODA over the same period, and it is directed toward very different targets. Non-OECD financiers have been active primarily in countries exporting oil (Angola, Nigeria, and Sudan) or other valuable minerals. Their involvement in African transport infrastructure has predominantly consisted of Chinese support for railway development and Arab support for roads. Financing: Filling the Gaps 339 Between 2001 and 2007, China provided financing commitments of around $4 billion to the African rail sector. The financing is for rehabilita- tion of more than 1,350 kilometers of existing railway lines and construc- tion of more than 1,600 kilometers of new railroad. (For a perspective on these figures, the entire African railroad network amounts to around 50,000 kilometers.) The largest deals have been in Nigeria, Gabon, Mauritania, and Angola, although the Nigerian deal has been delayed.2 Chinese financing in Gabon and Mauritania will facilitate export of iron and phosphate deposits, respectively. In Angola, part of a $2 billion credit is being invested in the refurbishment of the Benguela railway and the rehabilitation of the railway between the port of Namibe and city of Menogue. By contrast, the cumulative value of Chinese financing for the road sector was only $600 million for the period 2001­07. Arab development institutions committed much more--about $1.8 billion over the same period--to financing African roads (Briceño-Garmendia, Smits, and Foster 2009). China's official economic assistance quadrupled between 2001 and 2005, reaching more than 35 African countries. Most of the assistance has gone to resource-rich countries, some of it in the form of barter arrangements under the "Angola mode."3 This south-south cooperation builds on economic complementarities: China has a strategic interest in Africa's natural resources, while Africa harnesses China's construction capabilities, thus helping African countries to develop their economic infrastructure. The implementation processes for ODA and non-OECD finance are completely different. While ODA is channeled through the government budget, China tends to directly execute its own financing, often with associated imports of human resources. Although this approach raises significant challenges, such as ensuring that the recipient benefits from technical assistance in project implementation, it also circumvents some of the capital budget execution problems typically associated with public investment. Non-OECD finance also raises questions about sustainability. Non- OECD financiers from China, India, and the Persian Gulf states put their resources behind sectors, countries, and circumstances aligned with their national business interests. They offer realistic financing options for power and transport particularly for postconflict countries with natural resources. But nongovernmental organizations are voicing concerns about the social and environmental impact of these projects. And because 340 Africa's Transport Infrastructure non-OECD financiers rarely offer operational, institutional, or policy assistance along with their funding, it is unclear whether the new assets are sustainable. How the economic downturn will affect non-OECD finance is diffi- cult to predict. Such aid, funded by domestic taxpayers in the donor countries, may be particularly vulnerable to budgetary cutbacks. The downturn in global commodity prices may also inhibit Chinese infra- structure finance linked to natural resource development. Private Investors: Over the Hill? All values reported in this section exclude royalty payments to govern- ments for transport infrastructure, which--although valuable from a fis- cal perspective--do not contribute to the creation of new transport assets. Since the late 1990s, private investment commitments in African trans- port infrastructure have surged, from $900 million in 1997 to $4.6 billion in 2006. Accounting for project implementation cycles, this translates into an average annual disbursement of $1.1 billion a year, or 0.16 per- cent of GDP (see table 8.1). These disbursements are very similar in mag- nitude to those received from non-OECD financiers. For infrastructure as a whole, Africa's resource-rich countries have been receiving the largest volume of private participation. Low-income countries--including fragile states--are capturing annual average flows of well over 1 percent of GDP. Relative to their GDP, however, Africa's middle-income countries have done less well. The picture for transport infrastructure is somewhat different. Private capital flows to the African transport sector have been volatile over time (figure 8.11, panel a). Occasional spikes have been driven by the financial closure of a handful of large projects, such as the N3 toll road project in South Africa in 1999 (worth a total of $600 million); the slew of con- tainer terminal concessions at Nigerian ports in 2005 (worth a total of $700 million); and the Gauteng light rail concession in South Africa in 2006 (worth a total of $3.8 billion). Aside from these megaprojects, the average annual private capital flow to African transport infrastructure during the 2000s has been no more than $300 million. About 60 percent of private finance for African transport has gone to railways, accounting for around $5.2 billion of cumulative commitments. A further 24 percent has gone to toll road projects, amounting to cumulative commitments of $2.1 billion. The remaining 16 percent, or $1.3 billion, has been spent on seaports (figure 8.11, panel b). Private par- ticipation in African transport infrastructure is almost invariably through Financing: Filling the Gaps 341 Figure 8.11 Overview of Private Commitments to African Transport Infrastructure a. Over time (US$ millions) 5,000 4,500 4,000 3,500 commitment 3,000 2,500 2,000 1,500 1,000 500 0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 year b. By subsector (US$ millions) c. By country (US$ millions) airports others 1,017 Côte d'lvoire seaports 381 1,334 Uganda 400 Kenya rail South Africa 453 roads 5,155 4,981 2,057 Mozambique 708 Nigeria 932 Source: PPIAF 2008. Note: Cumulative investment commitments from 1990 to 2007, not disbursements. Chart is comprehensive; all countries excluded do not have any private investment in transport infrastructure. concession contracts. As noted in previous chapters, a very limited portion of Africa's road network meets the minimum traffic thresholds required to support toll road concessions. Rail concessions are more numerous, but investment commitments--especially realized investments--have fallen well short of requirements and expectations because of the limited traf- fic volumes on the lines and the constraints on tariffs imposed by inter- modal competition. Seaport transactions have primarily taken the form of container terminal concessions, which are becoming increasingly com- mon around the region. 342 Africa's Transport Infrastructure More than half of the total private investment in African transport infrastructure has gone to South Africa, which captured cumulative commitments of almost $5 billion between 1990 and 2007 (figure 8.11, panel c). South Africa captured about 70 percent of private investment in African railways and roads but hardly any of the private investment in ports and airports. The next closest country is Nigeria, which captured $900 million of cumulative commitments over the same period, concen- trated exclusively in ports and airports, with no private investment in roads or railroads. Eight countries have captured between $100 million and $700 million of cumulative commitments. In descending order, these countries are Mozambique, Kenya, Uganda, Côte d'Ivoire, Gabon, Tanzania, Cameroon, and Zimbabwe. In most cases, the bulk of these resources has gone to railroads, with the exception of one sizable toll road investment in Mozambique. Another 12 countries have captured modest amounts of private investment for transport, averaging around $40 million in cumulative commitments per country. There have been no toll road or airport investments in any of these countries--financing has been divided evenly between railroad and seaport projects. Countries such as Burkina Faso, Madagascar, Malawi, Mali, Senegal, and Zambia have also achieved significant railroad transactions, and Angola, Equatorial Guinea, Ghana, Madagascar, Mauritius, and Sudan have ben- efited from significant seaport transactions. The global financial crisis is likely to affect private capital flows even more than official flows. In the aftermath of the 1997 Asian financial cri- sis, private participation in developing countries fell by about half over a period of five years. Meanwhile, existing arrangements in Africa are com- ing under stress as parties encounter difficulties refinancing short- and medium-term debt. Local Sources of Finance: Possible in the Medium Term Local capital markets are a major source of infrastructure finance in South Africa, but this is not yet true elsewhere. Local infrastructure finance consists primarily of commercial bank lending, some corporate bond and stock exchange issues, and, more recently, the participation of a small but growing number of institutional investors. Along with information and communication technology, transport is reporting higher volumes of finance from local capital markets than are most other infrastructure sectors, although the absolute volumes remain small. The outstanding stock of finance for transport infrastructure on local capital markets was $17.1 million for South Africa and $6.2 million Financing: Filling the Gaps 343 for the remainder of Africa. These volumes are low when compared with an annual investment requirement of $8.8 billion for the sector in Africa (note that the total financing requirement of $19.2 billion includes main- tenance as well as new investment). The stock of outstanding bank loans to the transport and communica- tion sector was $8.5 billion at the end of 2006 (unfortunately, volumes for the two sectors cannot be disaggregated). South Africa accounted for about $5.0 billion of this total, with the rest of Africa making up the remaining $3.5 billion. This total represents only 2.9 percent of outstand- ing bank loans throughout Africa (Irving and Manroth 2009). As well as being limited in size, bank lending tends to be short in tenor for all but the most select bank clients, reflecting the predominantly short-term nature of banks' deposits and other liabilities. The longest maturities available--around 20 years--were found only in Ghana, Lesotho, Namibia, South Africa, Uganda, and Zambia. Eight other countries reported maximum loan maturities of 10 years or more. Even where 20-year terms are reportedly available, they may not be affordable for infrastruc- ture. In Ghana and Zambia, for example, average lending rates exceed 20 percent. Very few infrastructure projects generate sufficient revenues to achieve that rate of return. For most African countries, local banking systems are too small and too constrained by structural impediments, such as the lack of credits of ade- quate maturity, to assemble funds for infrastructural development. Syndicated lending to infrastructure projects with the participation of local banks, which has increased in recent years, may hold more potential. The volume of syndicated loans to infrastructure borrowers rose steeply from $600 million in 2000 to $6.3 billion in 2006, with 80 percent of this amount concentrated in South Africa (Irving and Manroth 2009). But there is so far only one example of a syndicated loan to the transport infrastructure sector in South Africa--the $475 million rand-denominated loan by three South African banks to Trans-African Concessions for con- struction of the N4 toll road. The second tranche of this loan had a matu- rity of 20 years. In the past decade, governments in the region have extended the matu- rity profile of their securities issues in an effort to establish a benchmark against which corporate bonds can be priced. With the exception of South Africa, however, such corporate bond markets remain small and illiquid. At 13.0 percent of GDP, South Africa's corporate bond market is by far the largest in the region, with $33.8 billion in issues outstanding at the end of 2006, followed by Namibia's, with $457 million (7.1 percent 344 Africa's Transport Infrastructure of GDP). Outside South Africa, the few countries that had corporate bonds listed on their national or regional securities exchange at the end of 2006 had only a handful of such listings, and the amounts issued were small. Only $1.1 billion of corporate bonds issued by transport infrastruc- ture providers was outstanding at the close of 2006 (table 8.8). The bulk of this--around $772 million--was issued by the South African National Roads Agency, and a further $298 million related to road financing in Namibia. Aside from the corporate bonds issued by these middle-income countries, the only other case found was a small issue of $62 million for the Port of Dakar in Senegal. The maturities reported on the transactions outside South Africa ranged from 6 to 10 years. The region's stock exchanges have played a more significant role in the transport sector, raising a total of $13.6 billion of capital--more than 80 percent in South Africa and the remainder in a handful of countries including Côte d'Ivoire, Nigeria, and Sudan. Institutional investors, including pension funds and insurance compa- nies, could potentially become an important source of financing in the Table 8.8 Outstanding Financing Stock for Transport Infrastructure as of 2006 Share Share of total Bank Corporate Equity of total infrastructure loans bonds issues Total stock (%) stock (%) South Africa (US$ millions) 5,011 772 11,269 17,052 73 28 Middle-income countries (excluding South Africa) (US$ millions) 142 298 -- 441 2 81 Resource-rich countries (US$ millions) 1,375 -- 87 1,462 6 56 Low-income, nonfragile countries (US$ millions) 1,728 62 2,173 3,963 17 54 Low-income, fragile countries (US$ millions) 278 -- 69 346 1 73 Total (US$ millions) 8,534 1,133 13,598 23,265 100 32 Share of total stock (%) 37 5 58 100 Share of total infrastructure stock (%) 12 2 19 32 Source: Adapted from Irving and Manroth 2009. Note: -- = not available. The stock includes bank loans, corporate bonds, and equity issues. The stock level reported under "transport" may be an overestimate because many countries report this category together with elements of communications and storage, and some countries together with electricity and water. Table is based on data from the following 22 countries: Cape Verde, Lesotho, and Namibia (middle-income); Nigeria, Sudan, and Zambia (resource-rich); and Côte d'Ivoire, Benin, Burkina Faso, the Democratic Republic of Congo, Ethiopia, Ghana, Kenya, Madagascar, Malawi, Mozambique, Niger, Rwanda, Senegal, Tanzania, and Uganda (low-income). Financing: Filling the Gaps 345 future, with more than $90 billion in assets accumulated in the former and more than $180 billion held by the latter. But as of today, less than 1 percent of those assets are invested in infrastructure, and even if that percentage were increased, it is likely that telecommunications would benefit much more than transport infrastructure. Harnessing the significant potential of local capital markets, particu- larly local bond markets, to finance infrastructure is thus contingent on the development of these markets as well as on further reforms to deepen the local institutional investor base. Well-functioning and appropriately regulated local institutional investors (pension funds and insurance com- panies) would be natural sources of long-term financing for infrastructure because their liabilities are well matched to the longer terms of infrastruc- ture projects. Private pension providers have begun to emerge as a possi- ble source of infrastructure financing with the shift from defined-benefit to defined-contribution schemes, which are viewed as less costly, more transparent, and easier to manage.4 Moreover, local institutional investors are also diversifying their portfolios, making infrastructure investments more attractive. With regional integration, financial markets could achieve greater scale and liquidity. More cross-border intraregional listings--of both corporate bonds and equity issues--and more cross-border intraregional invest- ment (particularly by local institutional investors) could help overcome national capital markets' impediments of small size, illiquidity, and inad- equate market infrastructure. They could also facilitate the ability of companies and governments to raise financing for infrastructure. So far, this intraregional approach to raising infrastructure financing remains largely untested. One new initiative is the Pan-African Infrastructure Development Fund, a 15-year regional fund that in its first round in 2007 raised $625 million for commercially viable infrastructure projects in Africa, including funds from Ghanaian and South African institutional investors. Costs of Capital and Sources of Finance Each source of infrastructure financing has a different associated cost of capital (figure 8.12). For public funds, raising taxes is not a costless exer- cise. Each dollar raised and spent by an African government has a social value premium (or marginal cost of public funds) of almost 20 percent, reflecting the incidence of that tax on the society's welfare (caused by changes in consumption patterns and administrative costs, among other things). To allow ready comparisons across financing sources, this study 346 Africa's Transport Infrastructure Figure 8.12 Costs of Capital by Funding Source public 1.17 India 0.91 China 0.87 source Arabs 0.65 ODA 0.51 IDA 0.33 grants 0.00 0 0.20 0.40 0.60 0.80 1.00 1.20 present value of $1 raised from different sources Source: Average marginal cost of public funds as estimated by Warlters and Auriol (2005); cost of equity for private sector as in Estache and Pinglo (2004) and Sirtaine and others (2005); authors' calculations. standardized the financial terms as the present value of a dollar raised through each of the different sources. In doing so, it recognized that all loans must ultimately be repaid with tax dollars, each of which attracts the 20 percent cost premium. Wide variation exists in lending terms. The most concessional IDA loans charge zero interest (0.75 percent service charge) with a 10-year grace period. India, China, and the Gulf states charge 4 percent, 3.6 per- cent, and 1.5 percent interest, respectively, with a 4-year grace period. The cost of non-OECD financing is somewhere between that of pub- lic funds and ODA. The "subsidy element" for Indian and Chinese funds is about 25 percent and for the Arab funds about 50 percent.5 ODA typ- ically provides a subsidy element of 60 percent; this rises to 75 percent for IDA resources. In addition to differences in the cost of capital, sources of financing differ in their transaction costs, which may offset or accentu- ate some of the differences. The Most Promising Ways to Increase Funds What are the best ways of increasing the availability of funds for infra- structure development? The way to start is clearly to get the most from existing budget envelopes, which can provide up to $2.4 billion a year of additional resources internally. For many countries, this would be enough to close the funding gap. But for a number of others--particularly the fragile states--a significant gap would remain even if all inefficiencies Financing: Filling the Gaps 347 were eliminated. Before the 2008 financial crisis, the prospects for reducing--if not closing--this gap appeared reasonably good. Resource royalties were at record highs, and resource-rich countries could use nat- ural resource savings accounts to provide financing for infrastructure (if macroeconomic conditions allowed). All sources of external finance were buoyant and promising further growth. As a consequence of the crisis, all sources of infrastructure financing in Africa may have less to offer rather than more, and the funding gap may widen further. It is worth noting, however, that the impact of the crisis has been less pronounced in Africa than elsewhere. In its World Economic Outlook issued early in 2010, the International Monetary Fund estimated that while economic growth in Africa had fallen from 5.2 percent in 2008 to 1.9 percent in 2009, it was expected to return to 4.3 percent in 2010 and rise further to 5.3 percent in 2011 (IMF 2010). The Residual Funding Gap The funding story is not entirely bleak. Assuming that current levels of ODA are unchanged and all efficiency gains are realized, 10 out of the 24 countries have more than enough funding to cover the estimated needs for the transport sector (figure 8.13). Even excluding ODA, four countries (Botswana, Malawi, Cape Verde, and the Republic of Congo) show small surpluses. But when the efficiency improvements are not allowed for, the picture is far less encouraging, with no country in the sample showing a surplus of financing over estimated needs (figure 8.14). Even when ODA is excluded and no allowance is made for potential efficiency gains, the funding gaps for some middle-income countries (South Africa and Cape Verde) and resource-rich states (Nigeria) amount to a very small proportion of GDP. Furthermore, addressing all the inef- ficiencies described above and capturing the user benefits for the public budget would be more than enough to close the overall funding gap for transport infrastructure at the aggregate regional level. This would also be the case in aggregate for the middle-income, resource-rich, and low- income, nonfragile states. For several reasons, however, this is no grounds for complacency. First, the estimated funding gaps for the transport sector (see figure 8.13) are very significant for half a dozen countries. For the Democratic Republic of Congo, the gap is close to 20 percent of GDP (or $1.4 billion), while Chad has a transport funding gap of more than 4 percent of GDP. Mali, 348 Africa's Transport Infrastructure Figure 8.13 Transport Infrastructure Financing Gap, Assuming All Efficiency Gains Congo, Dem. Rep. Chad Mali Niger Tanzania Madagascar Uganda Côte d' Ivoire Mozambique Cameroon Senegal Benin South Africa Nigeria Rwanda Ethiopia Zambia Kenya Ghana Lesotho Namibia Botswana Malawi Cape Verde Congo, Rep. ­2 ­1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 percentage of GDP gap gap without ODA and non-OECD funding Source: Adapted from Briceño-Garmendia, Smits, and Foster 2009. Niger, Tanzania, and Madagascar all have shortfalls of around 1 percent of GDP. As a group, the low-income, fragile states have a funding gap of $2.5 billion per year, representing more than 6 percent of aggregate GDP. About two-thirds of this funding gap relates to capital investment, with the remaining amount relating to maintenance spending. In addition, the benefit from increased user charges is actually a trans- fer from users to government, not a real efficiency gain per se. The actual economic efficiency gain--the largest category of potential efficiency gains--arises in the form of operating cost savings to users, greater than, Financing: Filling the Gaps 349 Figure 8.14 Transport Infrastructure Financing Gap, Excluding Any Efficiency Gains Congo, Dem. Rep. Chad Mali Niger Tanzania Madagascar Uganda Côte d'lvoire Mozambique Cameroon Senegal Benin South Africa Nigeria Rwanda Ghana Ethiopia Lesotho Malawi Zambia Kenya Namibia Congo, Rep. Botswana Cape Verde 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 percentage of GDP gap gap without ODA and non-OECD funding Source: Briceño-Garmendia, Smits, and Foster 2009. Note: Based on annualized averages for 2001­06. but resulting from, increased maintenance expenditures by the road agen- cies. Hence, converting the agency gain into real efficiency gains would be contingent on devising methods of recapturing some of the benefit to fund the investment. That puts the focus back on the need for strong and consistent policy making. While it may be unrealistic to expect that all the inefficiencies can be eliminated, even halving them would make a substantial contribution to the African transport sector. 350 Africa's Transport Infrastructure The Way Forward The estimated cost of meeting Africa's transport sector spending needs amounts to $19.2 billion a year. At first glance, this does not appear too far above current transport sector spending (including ODA) of $16.3 billion a year. Moreover, in aggregate, it would appear that 80 percent of the gap between spending needs and current spending could be eliminated by cap- turing an estimated $2.4 billion a year in efficiency gains, mainly in the road sector. The inefficiencies in question arise from underexecution of capital budgets ($1.3 billion a year), underrecovery of road user costs ($600 million a year), and undercollection of levies ($500 million a year). Even this seemingly positive picture leaves no room for complacency, for three reasons. First, the efficiency gains may not be easily achieved. The analysis underscores the importance of completing the reform agenda to ensure adequate funding for and spending on road maintenance, as well as to improve the effectiveness of project appraisal and procurement processes across the institutions responsible for implementing public investment programs (such as the road fund and road agency administrations dis- cussed in chapter 2). Second, the analysis depends on recent levels of ODA and other sources of external finance being maintained. Overall, these sources account for nearly one-quarter of transport infrastructure expenditures, and for the low-income, fragile states, they account for more than one- half. While statements of intent from the funding sources are quite promising, it still remains to be seen how well they survive the global financial crisis. Finally, the analysis does not apply to all countries. The extreme cases are the low-income, fragile states, two of which would have a funding gap of $2.5 billion a year even if all efficiency gains were fully captured. Raising more funds for transport infrastructure--particularly road infra- structure--in low-income, fragile states will be challenging. Historically, the main sources of financing have been public budgets and ODA, both of which are likely to suffer as a result of the financial crisis. The poten- tial of the private sector to contribute to road finance is relatively small, while non-OECD finance has tended to go preferentially to the railroad sector. Closing the gap for transport finance in low-income, fragile states will therefore likely entail delaying the achievement of targets or opting for lower-cost technologies and standards. A number of priorities for action can be identified. Financing: Filling the Gaps 351 Priority 1. Consolidating commitment to adequate and efficient maintenance Ensuring that maintenance is performed in an adequate and timely man- ner is the first priority. To this end, the following contributory actions are required: · Further strengthening of road fund administrations and procedures to permit multiannual contracts · Continuation of efforts to implement road works more effectively through quasi-independent road agencies. Priority 2. Ensuring better budget execution ratios Poor budget execution results in transport infrastructure expenditures falling below the budget allocations. A number of steps can be taken to improve this situation: · Better consultation among sector ministries, finance ministries, and donors on the balance between needs and absorptive capacity · Strict adherence to timetables for budget preparation · Establishment of medium-term programs for each subsector, taking into account absorptive capacity as well as needs · Maintenance of a pool of prepared projects so that any variations in budget availability can be accommodated. Priority 3. Improving the accuracy of needs estimates on a country-by-country basis The model used in this study is essentially a "broad brush" model. While it is available for further development and country application, it needs to be refined as follows: · More comprehensive in its coverage of all of the component elements of expenditure · Calibrated to local costs and conditions in each country. Priority 4. Sustaining ODA Many countries are heavily dependent on ODA for their investment expenditures on transport infrastructure. That support was likely to only increase before the ongoing world financial crisis. Now, if commit- ments are not maintained, there is a danger that some countries will divert even more of their domestic funding from maintenance to 352 Africa's Transport Infrastructure investment. The following steps may help to make sure that ODA is maintained: · Further shifts of ODA to sector budget support, contingent on ade- quate allocations and execution of maintenance programs · More commitment--on the part of both donors and recipients--to development policies and programs under the New Partnership for Africa's Development initiative. Priority 5. Making infrastructure finance more attractive to the private sector Concessioning has been used as a means of attracting private sector par- ticipation in infrastructure finance and management, particularly in the rail and ports sectors. But in the rail sector, this has not yet proved to be a sustainable source of capital funding. To make capital finance more attractive to the private sector, governments can commit to the following reforms of concessioning arrangements: · Development of new railway concession models under which govern- ments continue to participate in capital funding when major expan- sion of rehabilitation programs is required · Firm contractual commitment to the adequate and timely compensa- tion of public service obligations, particularly the maintenance of pas- senger services in rail concession contracts. Notes 1. Cecilia Briceño-Garmendia and Nataliya Pushak are the authors of the main source document for this chapter. William Butterfield, Chuan Chen, Vivien Foster, Jacqueline Irving, Astrid Manroth, Afua Sarkodie, and Karlis Smits also contributed. Rodrigo Archondo-Callao and Alberto Nogales performed the calculations based on the Highway Design and Maintenance Model 4 (HDM4) and the Road Network Evaluation Tool (RONET). 2. Although first announced in late 2006, the deal had not been finalized by January 2010, when it was announced that the first drawdown of the $500 million loan for the Kano-Lagos railway rehabilitation was imminent. 3. The Angola mode was devised to enable African nations to pay for infrastruc- ture with natural resources. In a single transaction, China bundles development- type assistance with commercial-type trade finance. A Chinese resource company makes repayments in exchange for oil or mineral rights. The China Financing: Filling the Gaps 353 Export-Import Bank acts as a broker, receiving money for the sale and paying the contractor for providing the infrastructure. This arrangement safeguards against currency inconvertibility, political instability, and expropriation. 4. For a defined-contribution pension scheme, under which pensions are based solely on the financial performance of saved contributions, infrastructure bonds may be a relatively attractive asset because of their long-term nature and the fact that they may give slightly better returns than other government bonds. 5. The subsidy element is the percentage difference between the present value of the loan and interest repayments of a loan that is made on standard market terms, and the present value of the loan and interest repayments of the loan in question. References Briceño-Garmendia, C., K. Smits, and V. Foster. 2009. "Financing Public Infrastructure in Sub-Saharan Africa: Patterns, Issues, and Options." Africa Infrastructure Country Diagnostic Background Paper 15, World Bank, Washington, DC. Calderón, C., and L. Servén. 2004. "Trends in Infrastructure in Latin America, 1980­2001." Policy Research Working Paper 3401, World Bank, Washington, DC. Carruthers, R., R. R. Krishnamani, and S. Murray. 2009. "Improving Connectivity: Investing in Transport Infrastructure in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 7, World Bank, Washington, DC. Estache, Antonio, and Maria Elena Pinglo. 2004. "Are Returns to Private Infrastructure in Developing Countries Consistent with Risks Since the Asian Crisis?" Policy Research Working Paper 3373, World Bank, Washington, DC. Foster, V., and C. Briceño-Garmendia, eds. 2009. Africa's Infrastructure: A Time for Transition. Paris: Agence Française de Développement; Washington, DC: World Bank. Foster, V., W. Butterfield, C. Chen, and N. Pushak. 2008. "Building Bridges: China's Growing Role as Infrastructure Financier for Sub-Saharan Africa." Trends and Policy Options 5, Public-Private Infrastructure Advisory Facility, World Bank, Washington, DC. Gwilliam, K., V. Foster, R. Archondo-Callao, C. Briceño-Garmendia, A. Nogales, and K. Sethi. 2009. "The Burden of Maintenance: Roads in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 14, World Bank, Washington, DC. IMF (International Monetary Fund). 2009. The State of Public Finances: Outlook and Medium-Term Policies After the 2008 Crisis. Washington, DC: IMF. ------. 2010. World Economic Outlook. Washington, DC: IMF. 354 Africa's Transport Infrastructure Irving, J., and A. Manroth. 2009. "Local Sources of Financing for Infrastructure in Africa: A Cross-Country Analysis." Policy Research Working Paper 4878, World Bank, Washington, DC. Lall, R., R. Anand, and A. Rastogi. 2009. "Developing Physical Infrastructure: A Comparative Perspective on the Experience of China and India." Unpublished paper, India Development Finance Corporation, Mumbai. ODI (Overseas Development Institute). 2009. A Development Charter for the G-20. London: ODI. PPIAF (Public-Private Infrastructure Advisory Facility). 2008. Private Participa- tion in Infrastructure Project Database. http://ppi.worldbank.org/. Reinikka, R., and J. Svensson. 2002. "Explaining Leakage of Public Funds." Discussion Paper 3227, Centre for Economic Policy Research, London. Sirtaine, S., M. E. Pinglo, J. L. Guasch, and V. Foster. 2005. "How Profitable Are Infrastructure Concessions in Latin America? Empirical Evidence and Regulatory Implications." Trends and Policy Options 2, Public-Private Infrastructure Advisory Facility, World Bank, Washington, DC. Warlters, M., and E. Auriol. 2005. "The Marginal Cost of Public Funds in Africa." Policy Research Working Paper 3679, World Bank, Washington, DC. World Bank. 2009. Global Economic Prospects. Washington, DC: World Bank. CHAPTER 9 Governance: The Key to Progress In each of the preceding chapters, the same message has emerged: inade- quate infrastructure goes only part way toward explaining the poor per- formance of the transport sector in Africa. On the one hand, existing infrastructure has been used inefficiently, increasing investment needs and subsequent fiscal demands; on the other hand, institutional and pol- icy deficiencies continue to mitigate the effectiveness of new investment. Given the history of clientelism and patronage-based decision making in Africa, it is recognized that effective governance is critical if sector devel- opment is to have a positive effect. Even Julius Nyerere, while rejecting what he considers the neocolonialist approach to the subject adopted by many aid agencies, accepts that many of Africa's problems arise from bad governance (Nyerere 1998). This chapter examines the impact of gover- nance quality on transport infrastructure in Africa.1 The "Washington consensus" that emerged during the 1990s is based on a belief that good economic policy is essential to growth (Burnside and Dollar 1997). Some observers have gone so far as to equate good gover- nance solely with good economic policy, and have tried to demonstrate, empirically, which economic policies are most conducive to growth (Osborne 2004). 355 356 Africa's Transport Infrastructure Rather more broadly, governance has been defined as the means by which government is exercised (Kaufmann, Kraay, and Mastruzzi 2010). The Worldwide Governance Indicators (WGI) project of the World Bank defines governance as the traditions and institutions by which authority in a country is exercised. This considers the process by which governments are selected, monitored and replaced; the capacity of the government to effectively formulate and implement sound policies; and the respect of citizens and the state of the institutions that govern economic and social interactions among them. (World Bank 2010b) This chapter uses this broad interpretation of the nature and role of governance. The Context of National Governance The most comprehensive attempt to assess the quality of governance worldwide is contained in the WGI report, produced annually by the World Bank. The WGI database uses the perceptions of multiple groups to develop an indicator set. The indicators relate to nations' overall economic levels and not to any specific sector. Each of the 212 countries in the database is given a mark between 2.5 and 2.5 on each of six criteria. Data for the estimation of the values are derived from a wide range of sources, details of which are set out in the data- base (World Bank 2010b).2 The following criteria for governance quality are adopted: · Voice and accountability · Political stability and absence of violence · Government effectiveness · Regulatory quality · Rule of law · Control of corruption. The database shows governance in Africa in a poor light. While there are bad outliers in several regions (for example, Myanmar in East Asia and the Pacific, and Afghanistan in South Asia), and while Africa is not the worst region in the world (Central Asia, mainly comprised of the former Soviet Union countries, holds this place), the average score for Africa is the second worst, comparable with South Asia. Many African countries fall in the lowest quartile of rankings on all dimensions, and most fall in Governance: The Key to Progress 357 the lowest half. Only Namibia and Botswana appear in the top half of the rankings on all six dimensions, though South Africa and Ghana come close. Any discussion of the transport sector thus needs to be seen against the backdrop of what most agree is a generally poor level of governance across the region. In turning more specifically to the transport sector, we concentrate on three issues relevant to the WGI definition of governance: · The significance of national and regional traditions for sector operations · The appropriateness and effectiveness of the institutions in the sector · The human resource capacity of the country in question. Traditions and Attitudes The first aspect of governance emerging from the WGI definition con- cerns traditions and attitudes. Countries may have longstanding, entrenched attitudes that counter explicitly stated government objectives or that are unacceptable to the governed. This section considers some of the attitudes and traditions that have been particularly damaging in the context of transport infrastructure, including corruption, the unrealistic expectations of the private sector, state capture by elites, and the absence of a safety culture. Another harmful tradition, aid dependence, is dis- cussed later in the chapter. Corruption The postcolonial history of Africa is an almost unrelieved story of corrup- tion and the decay of governance, which has been well documented else- where (Meredith 2005). That is not the subject of this book, but it must be noted at the outset, as many of the failings in the transport sector are attributable to the attitudes toward power and governance that have affected all sectors. While there have been improvements in some coun- tries, the recently released Corruption Perceptions Index (Transparency International 2009) finds 10 African countries in the bottom decile (with Somalia at the very bottom of the list). Of the 47 African countries reviewed, 31 scored less than 3 out of 10, "indicating that corruption is perceived as rampant." Without a more fundamental return to good gov- ernance, many of the reforms recommended for the transport sector will be either thwarted or perverted. Various forms of corruption existed prior to decolonization. In West Africa, in particular, there was a long tradition of "dash"--gifts made in 358 Africa's Transport Infrastructure recognition of services rendered by public officials and others. What independence did was to enshrine this practice as an institutionalized element of everyday life. Because the new rulers were themselves effec- tively unconstrained, they were able to exploit their positions by rou- tinely selling government contracts and positions to those willing to pay for them. As a consequence, the practice of bribery spread from top to bottom, particularly in systems for tax collection, customs administra- tion, and policing (box 9.1). The conditions conducive to corruption--the combination of a monopoly and administrative discretion--frequently arise in the trans- port sector. The artificial creation of queues for licenses, customs clear- ance, and so on provides a platform for the extraction of bribes. The poor performance of the African economies in the development of modern logistics systems, discussed below, is largely attributable to the widespread practice of bribery. Sequeira and Djankov (2008) find that bribe pay- ments at ports in southern Africa, while varying by port and product, are generally high and frequent. Bribes can increase total shipping costs by up to 14 percent for a standard 20-foot container and the monthly salary of a port official by about 600 percent. Bribery thus produces a diversion effect as firms take the long way around to avoid the most corrupt port, and a congestion effect as this rerouting increases congestion and transport costs in the region. Further, though the cost of serving select corridors is the Box 9.1 The Kenyan Government's Purchase of Luxury Vehicles for Official Use A Transparency International (TI) report found that between January 2003 and September 2004, the government of Kenya spent at least $12.1 million on the purchase of luxury vehicles, largely for the personal use of senior government of- ficials. The Ministry of Roads and Public Works ranked second among ministries in spending on luxury vehicles. The ministry spent close to $840,000 on the pur- chase of one Mercedes E240, four Land Cruiser Prados, two heavy-duty Land Cruisers, three Mitsubishi Pajeros V76 GLX, and a Land Rover Freelander. None of these expenditures fell within the cost estimate ceiling for government-fleet pas- senger vehicles stated in the Government Financial Management Act of 2004. Source: Transparency International, Kenya and Kenya National Commission on Human Rights 2006. Governance: The Key to Progress 359 same, overland transport rates for firms shipping to the most corrupt port are 71 percent higher than rates charged to the least corrupt port because of an imbalance in cargo flows. Other important aspects of transport infrastructure and service develop- ment are similarly blighted by corruption. Major transport infrastructure projects typically require ministerial, or even government, approval, and are thus susceptible to influence at the highest level. "Gold-plated" schemes, such as newly constructed airports, are particularly remunerative to the government and thus distort investment priorities. The creation of large parastatal transport-service providers has also been judged by international financial institutions as subject to malign influence. As a consequence, after a period of unsuccessful lending to sup- port state-owned enterprises, the international financial institutions, led by the World Bank and the International Monetary Fund, began to emphasize funding tied to structural adjustment. Notable in this shift was an emphasis on the privatization of the inefficient parastatals to which two-thirds of the structural adjustment lending of the late 1980s was directed. Even this effort was thwarted in many cases, as strong govern- ments took the opportunity to undertake secret, noncompetitive deals with friends and supporters (often in the military) to further bolster their own control of power. Eventually even the World Bank, which had been a strong supporter of privatization programs, concluded that economic reforms were likely to be unsuccessful without political reform to ensure good governance (World Bank 1989). A number of the most notable transport infrastructure initiatives since the 1980s originated from this change of emphasis. Second-generation road funds were designed to depoliticize decision making on the allocation and administration of road maintenance funding (see chapter 2). The rail concessioning program in Africa, which followed from the successful experiences of Latin America, concentrated on ensuring a transparent process of competitive bidding as the basis for the transfer of management responsibility (see chapter 3). Similar emphasis on fair process is also to be found in the move toward airport service concessions (chapter 4) and landlord port structures (chapter 5). While some experience in these areas has been positive, the battle is not yet won; further suggestions for improv- ing transport governance are contained in the final chapter of this book. The Role of the State and the Private Sector Many of the early postcolonial governments were heavily influenced by Marxist philosophy and saw socialism as the preferred political model. But, 360 Africa's Transport Infrastructure curiously, only President Sekou Touré in Guinea undertook the wholesale nationalization of existing enterprises. For example, in many countries, the former colonial expatriate bus companies were allowed to continue operating--but subject to new and more stringent regulations that ulti- mately destroyed most of them. As the state took a more directive role in the economy, whether in encouraging industrialization (as in Ghana under President Nkrumah) or developing the agricultural base (as in Tanzania under President Nyerere), the creation of new state enterprises was seen as a natural instrument of state-sponsored development. Moreover, as the "lib- erators" sought to increase their hold on power, the appointment of man- agers of newly created state enterprises became an important instrument of patronage. Business management skill was rarely a criterion for selection, and many of the new enterprises were doomed to failure from the start. An analysis of the state's role in development has evolved over the past half century, and the debate over the state versus the market has shifted focus to the more fundamental crisis of state effectiveness. In some countries--for example, Sierra Leone--the crisis led to an outright collapse of the state. While state-dominated development has failed, it is now commonly accepted that development without an effective state is also impossible. The challenge for most developing countries is to build on the relative strengths of private markets while taking into account and improving the state's institutional capability. Building the requisite state capacity in Africa through institutional reform, sector reorganiza- tion, technical assistance, and policy advice has been supported widely by the development agencies over the last two decades. Africa has seen some success in the reform of institutions overall. For example, most recently, Rwanda has moved from 143rd place to 67th place in the rankings developed by the Doing Business Indicators (World Bank 2010a). But the record in general is quite mixed. The most common problem lies in the poor understanding and unrealistic expectations that governments have of private sector concessionaires or franchisees. Excessively stringent fare controls, imposed without compensation, are largely responsible for the decline of formal bus companies, whether pri- vately or publicly owned. Self-regulation by unsubsidized operators in both trucking (chapter 2) and informal minibus operations (chapter 6) has stabilized supply, but usually in the interests of operators rather than customers. Attempts to maintain loss-making rail passenger transport services through the terms of concession contracts (chapter 3) threaten to have a similarly disastrous effect in some cases. A better understanding of the impact of concession terms on the profitability of rail infrastructure Governance: The Key to Progress 361 will be necessary to attract significant private financing of new rail infra- structure. Ensuring that the costs and risks of rail infrastructure develop- ment are shared by the state and concessionaires may also be necessary. Two conclusions can be made regarding this aspect of governance. On the one hand, cross-subsidization within protected state enterprises has not solved the problem of serving thin transport markets in Africa. On the other hand, the mechanisms for mobilizing private sector capital and ini- tiative have been imperfect and require substantial improvement going forward. The Role of Elites, or State Capture Capture of the policy-making process is a subtle form of corruption. Private interests, working through or with politicians, influence the direc- tion and content of policies to favor their own activities and investments. For example, the government concession of resources such as mineral deposits or hardwood forests in exchange for investment in a major trans- port asset, such as an airport or port, may be an effective way of mobiliz- ing private finance. But it may also be a way for politicians to secure financial gain at the cost of public benefit. State capture is also seen in the distribution of high-level appoint- ments and the allocation of responsibilities and mandates related to the planning and funding of transport infrastructure, with current political party leadership often the most important agent. Other agents are the educated and wealthy who together make up the African power elite and present themselves as the legitimate representatives of the larger society. There is often substantial discretion in planning and funding decisions that are the mandate of top-level executives or elected officials. The lack of adequate consultation, and the absence of objective planning criteria or analysis, create opportunities for influencing the distribution of sub- stantial public resources and incurring substantial social damage. Since the 1980s, attempts have been made to separate operations from policy-making and regulatory functions in a number of transport sectors. It was expected that this decentralization of power would allow greater impartiality and transparency and, depending on legislative and contrac- tual mandates, provide stronger incentives and controls for accountabil- ity. The resulting structures were considered less vulnerable to state capture than a vertically and functionally integrated public sector agency. But the mixed performance of such unbundled structures in recent times offers important lessons. For an unbundled structure to perform successfully, there needs to be both adequate capacity in each of the new 362 Africa's Transport Infrastructure sector entities and effective market forces. Where the sector is small and professional capacity weak, or where corruption spans public and private sectors, unbundling may even increase the opportunities for corruption and governance failure. For example, under the early road fund models, corrupt officials were sometimes able to capture road boards that had been given additional autonomy over substantial funds. Under the more recent second-generation road funds, oversight has been made stronger and more transparent through the inclusion of road users on the boards; however, where appointments are subject to high-level political authori- zation, the process can still be compromised through the direct appoint- ment of corrupt players. Similarly, corruption has flourished in semiautonomous port authorities and services where the appointments have been political or oversight weak. Thus, sector restructuring will reduce corruption only to the extent that the governance environment and institutional capacity are improved. Absence of a Safety Culture The African transport sector has the worst safety record of all world regions. This is most apparent in the air, road, and urban transport sec- tors, though there are also problems to be addressed in the port and rail sectors. At first glance, the reasons for this poor safety record appear to differ substantially across sectors. In rail transport, the problems arise from low standards and poor maintenance of infrastructure. In road transport, there is a combination of poor infrastructure, inexperienced vehicle operation, and poor law enforcement. In air transport, the source of problems seems to be the lack of adequate training and supervision of flight crews, together with the inadequacy of the air traffic control systems. Urban transport suf- fers from inadequate separation of traffic from other activities as well as inexperienced operators and poor enforcement of regulations. On further inspection, several factors can be found in common across sectors. First, when there are other serious threats to life and limb-- including the perils of war, starvation, and pestilence--transport accidents become only one more source of danger and hence are not seen as a high- priority issue. Further, where governance is weak and enforcement cor- rupt, a comprehensive effort to come to grips with the problem is difficult in any sector. Few countries have established effective oversight; instead, there seems to be a pervasive resignation to transport accidents as an unavoidable peril of life. Ghana's National Transport Safety Council is an exception to the rule, and supports the general conclusion that it is Governance: The Key to Progress 363 possible to change the culture surrounding safety only by a major initia- tive sponsored by and supported at a high level of government. Institutions The institutional framework through which governance is achieved is taken to encompass the formal and informal institutions of a country--its bureaucracy, private sector, nongovernmental organizations, judiciary sys- tem, civil society, and so on--as well as its laws, the ways and means of enforcing these laws, the procedures in place for mediating conflict, and the sanctions enforced when laws are breached. National Transport Institutions Institutions play a central role in setting out a country's or sector's objec- tives and translating them into action. The existence of institutional capa- bilities for strategic thinking and long-term policy development and for actual implementation, evaluation, and control of policies is critical for growth. Implementing institutions determine how factors of production-- land, capital, and labor--are obtained, how they are transferred, and how they are used. Legal and regulatory institutions are essential for efficient markets as they structure both the individual's and the firm's incentives for innovation, production, and exchange. Impartial contract-enforcement procedures also provide incentives for the development of complex com- mercial agreements. Defined and transparent procedures enhance pre- dictability by restraining opportunism and reducing the arbitrary influence of elites. This greater predictability reduces the costs and increases the ben- efits of economic exchange. Good institutions thus perform a number of functions: (i) they decrease information asymmetries as they channel information about goods, participants, and market conditions; (ii) they reduce risk as they define and enforce property rights and contracts determining who gets what and when; and (iii) they restrict the actions of politicians and inter- est groups by making them accountable to citizens. National transport policy. Governments generally organize the transport sector in Africa through a ministry of transport mandated to develop and implement policy. Most African countries have a formal transport-policy statement, and many have a long-term investment program. Just over 60 percent have a long-term road investment program (SSATP 2007), and this tends to be newer than the sector policy statement. As more 364 Africa's Transport Infrastructure countries increase their dialogue with stakeholders on medium-term expenditure plans, particularly in the context of sector budget support, their investment programs are becoming more realistic. In some coun- tries, however, such programs have little or no effect, mainly because of a lack of resources and (in some cases) lack of will to give expression to their goals and objectives. Implementation of national transport policies is also typically weak and fragmented. In the case of roads, a ministry of construction as well as local governments may be involved in building, rehabilitation, and main- tenance. Governments can choose to carry out these functions on their own or contract out some or all of these functions to the private sector. In some countries (though this is not common in Africa), there is a sepa- rate railways ministry, and it is quite common for ports and airports to be handled by a trade-related ministry. The most common failing of this institutional structure is to leave sec- tor strategies uncoordinated. In particular, rail administrations frequently complain that their own problems stem, at least in part, from road users being undercharged. A similar argument is often applied to the issue of road vehicle overloading. While the research is inconclusive, much evi- dence suggests that if road users were properly charged for the use of road infrastructure, more funds would be allocated to road maintenance, which would drastically reduce vehicle-operating costs. Similarly, a proper analysis of the costs of overloading might lead to decisions to spend more on initial road construction to handle higher axle loads, funded out of user charges, with a consequent reduction rather than increase in the costs per tonne-kilometer of road freight. The problem with this alternative approach is that the costs of upgrading the whole network could be very large, while raising axle-load limits without first having done the strengthening might cause additional counterproductive deterioration of the unimproved sections. This is an area where some fur- ther analytic work, using the data collected by RONET (Road Network Evaluation Tool) in the Africa Infrastructure Country Diagnostic (AICD) program, is called for. Road transport. For road infrastructure, the main failure of governance has concerned the inadequacy of funding for maintenance. The reason for it was the failure of governments, and road agencies, to take into account the implications of their maintenance expenditure decisions on road trans- port operating costs. The institutional solution has been the establishment of second-generation road funds, with private sector representation on Governance: The Key to Progress 365 management boards. Road infrastructure conditions are generally better in countries with such funds than in those without them (see chapter 2). But current arrangements still do not meet the objectives of governments or the aspirations of users. Underfinancing of maintenance continues, to the tune of over $1 billion per year for the region as a whole, and users suffer losses of four or five times that amount as a consequence. The reasons for continued underfunding of maintenance are largely institutional. Many of the funds remain based on administrative decree rather than on law. Funds are still frequently channeled through the national treasury rather than being paid directly, and fuel levies are set by ministers of finance with macro considerations in mind rather than by a road fund board operating as a commercial asset manager. There is thus much to do to improve the design of road funds and to convince govern- ments to allow them to operate in a commercial manner. Moreover, allo- cation does not mean disbursement: one of the main problems in some countries has been that allocated funds were not disbursed. This applies to capital investment as well as to maintenance funding. Without contin- ued commitment to a road fund's operation as a user-based agency to secure a more businesslike treatment of road maintenance, simply having a road fund does little to improve road conditions. Commercialization has made even less progress at the implementation level than at the financial management level. Fewer than half of the ini- tial 24 AICD countries have set up quasi-commercial road authorities. Even where quasi-autonomous agencies have been established, they often lack a commercial environment and continue to act like government departments without a clear understanding of their mission, functions, and work programs. Devices such as performance-based maintenance contracting, made possible by the existence of a fund not tied to the budgetary cycle, are employed less in Africa than in other regions. Meanwhile, road transport operations are almost completely in private hands. Entry into the trucking industry, though formally requiring a license, is not very restrictive, and the industry is fragmented. But most countries in West and Central Africa have very strong truckers' unions, which, particularly at the major ports, control the allocation of traffic to members and effectively set rates. This cartelization of the road haulage industry, as well as that of urban bus systems, leads to inefficient operat- ing practices. There is substantial evidence of severe overloading of vehi- cles, which not only is dangerous but also damages roads. In terms of the WGI criteria, the failures lie both in the appropriateness of the regulatory regime and in the efficacy of government implementation. 366 Africa's Transport Infrastructure Rail transport. Nearly two-thirds of the countries in the AICD sample have concessioned their rail networks to private sector operators. In most cases this has improved their finances, but in very few has it generated any substantial amount of new investment financing. This result reflects both the fragile financial condition of many of the concessionaires and the con- tinuing failure of many governments to recognize the realities of current conditions. In particular, the obligation to maintain some passenger serv- ices, whether formerly subsidized as a public service obligation or not, is commonly a drain on the concessions. Some concessions have already been renegotiated, and it is certain that a new form of public-private part- nership, in which the government accepts more responsibility for infra- structure finance, will be necessary if further collapse is not to occur. By far the largest rail system of all, in South Africa, remains in govern- ment hands, though in two separate organizations. The Passenger Rail Authority of South Africa is responsible for passenger services, while Transnet Freight Rail, formerly Spoornet (a subsidiary of the Transnet conglomerate, which also provides pipeline port infrastructure and port operations), is responsible for all freight services. Transnet Freight Rail benefits from having two very profitable freight lines--one carrying nearly 70 million tons of coal exports to Richards Bay and another carry- ing about half that quantity of ore exports to Saldanha--that cross- subsidize its general freight activities. It is also not burdened by generally less-profitable passenger services. It is doubtful whether the very opaque nature of cross-subsidization, inherent in this organizational structure, leads to the best decisions on resource allocation. Air transport. Airports are generally operated by public corporations or by regional governments, though the larger ones increasingly concession specific services to the private sector. They are financed through charges to their customer airlines and service providers, which are ultimately passed on to individual passengers. In most cases, both types of charges are high by international standards. Some deficit financing is provided by their owners. Air traffic control and navigation services are usually the responsibility of a civil aviation authority. But revenue usually accrues to the government, with only a portion being retained by the authority. Air transport services are provided in most countries by a mixture of national and foreign companies, some private and some public. Through their ultimate ownership of national landing rights, governments can exercise effective veto control of the services provided, but they are rela- tively powerless to generate service provision. Attempts to do this through Governance: The Key to Progress 367 state-owned national flag carriers, with an expectation that these carriers would be able to cross-subsidize loss-making routes, have not been suc- cessful. While about half of the countries in the AICD country sample have a national flag carrier, most of these are small and weak (with the exception of those of Ethiopia, Kenya, and South Africa). Most intercon- tinental services are provided at a lower cost by large foreign operators. Port and maritime and inland waterway transport. In the port sector, many of the smaller countries have stuck to the traditional public service port-management model, though it is clear that the ports of the region that have moved either partly or completely to the landlord model per- form better. This is not just because the concessionaires bring infrastruc- ture finance but also because they bring best international practices to African port management. The shipping system is rather better, largely as a consequence of the disappearance of the old measures protecting national shipping set by the former Liner Code of the United Nations Conference on Trade and Development. This liberalization has reduced costs, but the limitations of market size and port capacity--both in terms of channel depth and quay facilities--still discourage direct service on the main global itineraries. Urban transport. Most cities have no transport authority to guide the development of their transport system; meanwhile, the few nascent authorities (for example, the Lagos Metropolitan Area Transport Authority) tend to be weak and poorly funded. The result is that there is usually little coordination between the policies governing public trans- port and roads and the necessary infrastructure. Though about half a dozen countries have some form of public sector bus company, only Anbessa in Addis Ababa is the dominant operator in its area. For the rest, service is predominantly provided by informal oper- ators subject to little or no public regulation, and hence managed by oper- ators' associations. In many countries, there is no effective institution to exercise quality control over the operations of informal service suppliers. Such reliance on self-regulation is usually of benefit to suppliers rather than to passengers. Regional and International Institutions Several of the issues identified in this book have proved difficult for gov- ernments to handle because they were in one way or another suprana- tional. Such issues include coordinating customs arrangements across land 368 Africa's Transport Infrastructure transit corridors, choosing optimum locations for container transshipment facilities, licensing air transport services, implementing better air-safety and navigation systems, and developing international rail networks. Because many of the countries are small, or landlocked, or both, problems of governance in these areas are critical. As part of a review of transport governance in Africa, it is therefore necessary to look at the various inter- national institutions of relevance to the sector. Those institutions fall into four categories. Some are Pan-African and some are regional. And in each of these categories, some are multisector and some are sector or subsector specific. This brief review attempts to identify the role that each institution plays and its relation to other insti- tutions, the factors that limit operational success, and the prospects for improved performance in the future. Pan-African institutions. The Pan-African institutions bring together 53 African countries with a vision of regional integration across many sec- tors. Since the early 1960s, the United Nations Economic Commission for Africa has encouraged African states to combine their economies into subregional markets that would ultimately form one Africa-wide eco- nomic union. The Organization of African Unity, established in 1963, envisioned an African Economic Community from its outset. Progress in that direction has been slow. The Abuja Treaty of June 1991, which came into force in May 1994, provides for the African Economic Community to be set up over a 35-year period. A first step is strengthening existing regional economic communities (RECs) and creating new ones where needed. Tariffs and other barriers to regional trade would be harmonized within each REC to establish at the REC level a free trade area and a cus- toms union. It was envisaged that eventually all RECs would be harmo- nized, and an African Central Bank and a single African currency established as part of an African Economic and Monetary Union. Efforts to achieve these Pan-African goals were revitalized by the Constitutive Act (2000) that created the African Union and confirmed the RECs as the building blocks of that union. The union also integrated the New Partnership for Africa's Development (in 2001), which included an Action Plan on Infrastructure (in 2002). Within this plan, the RECs have specific responsibilities for coordinating infrastructure development at a continental level and are backed by the financial and technical support of the African Development Bank. Thus the institutional framework of the key organizations is now in place for achieving the goals of the African Union. Governance: The Key to Progress 369 Implementation of the wider political and economic agenda is pro- gressing. The coverage of the RECs is comprehensive, and most have free trade areas and customs unions either fully in force or near completion, but there has been no progress yet toward their establishment at a continent-wide level. Still, the goal of free trade and deeper integration remains an overriding priority for the African Union. The significance of Africa's efforts to increase regional integration in the transport sector is that free trade would eliminate many of the imped- iments to movement across national borders that have kept transport costs high. But a review of the state of regional transport integration painted a rather discouraging picture (UNECA 2004). The review shows there is a need to refocus policies, agree on and maintain appropriate net- works, improve transit facilitation, increase the capacity of regional bod- ies, and mobilize much-needed investment. In the interim, the African Union has yet to exercise a strong influence on the transport sector, although coordination between the African Union Commission and the RECs is promoting the high-priority regional and cross-border projects of the New Partnership for Africa's Development. Regional economic communities. The RECs are now well established. A protocol on relations between the African Union and the RECs was adopted in July 2007. Yet many operational difficulties still exist. In some instances, the activities of individual RECs overlap, as for example in the relationship between the Southern African Development Community and the Common Market for Eastern and Southern Africa (COMESA). Several countries are members of more than one REC. Cooperation at the regional level is further complicated by the fact that the RECs are not progressing toward an economic union at the same pace or by means of similar procedures and processes. It is not even clear that all existing RECs have the same long-term continental integration in view, or that there is the political will within all the RECs to submit regional concerns to the vision of the union. The RECs have already had some significant impacts on the transport sector. The establishment of free trade areas and customs unions has facil- itated coordinated development of transport corridors. Successes have rested on the mutual interest of neighboring countries, as in the Maputo corridor, where South Africa and Mozambique have coordinated plans for road and port development with border-crossing facilities. Arguably the greatest influence of the RECs, however, has been in air transport. In particular, the West African Economic and Monetary Union 370 Africa's Transport Infrastructure and COMESA have been the focus for the liberalization of international air services provided for under the Yamoussoukro Decision of 1999. Though so far there has been little impact on the intercontinental or domestic markets, two-thirds of African air transport is now liberalized. REC actions in other transport sectors have been less successful. There is a case for regional agreement on regional transshipment to a hub port in both East Africa and West Africa. But national competition for this trade continues, and several ports of suboptimal scale remain in business. Regional associations have little authority over member countries that fail to implement regulations agreed on at the regional level. Moreover, RECs suffer from insufficient staff, inadequate budgets, and dissimilar and often complex decision-making processes. Pan-African Transport Associations. Pan-African transport associations exist in all subsectors. Many are offshoots of broader international feder- ations. Rail systems are overseen by the African Union of Railways, which has produced plans for Pan-African trunk networks. But there has been little progress, largely because the individual links in such networks would require major investments by national authorities with more pressing infrastructure needs to consider. In the road sector, the initiative for the development of an African highway network has come primarily from the United Nations Economic Commission for Africa rather than from a Pan- African road infrastructure body. In air transport, the African Civil Aviation Commission (AFCAC)--a specialized institution of the African Union--has yielded the initiative to liberalize air service to the RECs. Similarly, in the troubled field of air transport safety, despite the existence of the AFCAC, most progress is likely to be achieved at a subregional level. In urban transport, the African Union of Public Transport (known also by its French acronym, UATP) has lobbied for more sustainable arrangements for urban public transport in the region. Despite their pro- liferation, few of these associations have great influence. Real influence comes only when an association has power to intervene, as in the case of the application of airport safety standards by the International Civil Aviation Organization. Regional transport associations. Regional transport associations are also active in some sectors. The South African Railways Association, while ini- tially established as a lobby group for equal treatment of road and rail infrastructure financing, has acted effectively in the planning of regional rail corridors. The parallel Association of Southern African National Road Governance: The Key to Progress 371 Agencies attempts to improve coordination between countries in road development. Regional port management associations--the Port Manage- ment Association of Eastern and Southern Africa and the Port Manage- ment Association of West and Central Africa--attempt to coordinate and standardize port services in their regions, but have not been able to put together effective regional port development strategies. In air transport, the formation of the East African Civil Aviation Authority heralds an attempt to address the problem of staff recruitment, training, and reten- tion through regional action. Africa is thus not short of regional and international associations with a hand in transport. But few of these are secure or successful. It seems that international collaboration is most likely to occur either where an international supervisory body has some real leverage (as in air transport, with the International Air Transport Association and the International Civil Aviation Organization) or where there are a limited number of part- ners, all of whom can see the mutual benefits of combined action. This explains both the unsuccessful network plans and port concentration efforts and the essentially subregional nature of developments in the air transport and rail sectors. It is vital that Africa's relevant regional institu- tions improve their governance, coordination, and operations if transport infrastructure and services are to reduce the cost of doing business in Africa. Capacity The final requirement of good governance is the capacity to implement and oversee. There are two dimensions to this: financial and human resources. Financial Capacity Constraints and Aid Dependence The financial analysis outlined in chapter 8 showed that very few coun- tries are in a position to maintain existing infrastructure or to undertake the investments necessary for even modest improvements. It will there- fore be necessary for them either to settle for even lower standards or to continue for some years to depend on loan and grant funding to satisfy their aspirations. Aid dependence is particularly high in countries that have suffered from economic crises, civil wars, and political instability. For example, in 2006, aid covered about 50 percent of road expenditures in Senegal and almost 90 percent in Rwanda. Many of the other low-income countries in 372 Africa's Transport Infrastructure Africa also receive high levels of aid. Moss and Subramanian (2005) iden- tify 16 low-income countries in Africa where inflows of official develop- ment assistance are equivalent to at least half of total government expenditure. Of the 12 poor countries where official aid accounted for 75 percent or more of government expenditure, 10 were in Africa. The concentration of aid is even higher when reviewed at the sectoral level. Such high levels of aid--when steered by clear development agendas-- can be used to improve policy and planning capacity, establish strong institutions, and strengthen civil services, as the experiences of the Republic of Korea and Taiwan, China, make clear. Botswana shows that such processes can work in Africa. Yet for many African countries, while aid can release governments from binding revenue constraints and help strengthen domestic institutions in the short term, it may make it more difficult for good governance to develop if continued over the long term (Brautigam and Knack 2004). There are several reasons for this weakness in governance. States that can raise a substantial proportion of their revenues from the donor com- munity are less accountable to their citizens and have less incentive to invest in effective public institutions. Bureaucrats are rewarded for getting money from donors rather than focusing on core development functions and leading deep-rooted institutional changes. Most African countries have undertaken donor-recommended reforms in the transport sector, meeting requirements for continued aid. Such reforms are often legal measures that are easily accomplished on paper (such as the establishment of second-generation road funds). Regulatory reforms that require more fundamental institutional changes and challenge vested interests have been much slower to emerge. Furthermore, large amounts of aid can also block governance improve- ments through institutional destruction and the creation of adverse incen- tives (Brautigam and Knack 2004). High numbers of donor-funded projects and reform agendas--each of which requires oversight, follow- up, and reporting--impose a high administrative burden on governments' often low and limited (absorptive) capacity. Senior officials spend much time facilitating or supporting donor supervision visits. The already low capacity is further weakened by donor competition for scarce staff skills and the provision of technical assistance that substitutes for the govern- ment's own capacity. Because governments cannot possibly manage the large number of projects donors want to fund, donors have set up units with off-budget funding. Technical-assistance staff members in these units seldom transfer skills but do the work themselves, thus limiting the Governance: The Key to Progress 373 government's ability to learn through doing (Gwilliam 2007). In addition to international staff, local staff are needed; in many countries, trained people are scarce. Donors consequently bid up the price of capable staff, poaching from both the private sector and the government. This poaching weakens institutions as it leaves them depleted, creates resentment, and lowers the morale of those left behind. Within the road sector, the availability of donor funding for reconstruc- tion leads to a neglect of routine maintenance--and so the conditions of existing roads start to deteriorate. Apart from the additional costs imposed by lack of maintenance, such funding allocation also undermines the development of maintenance planning and budgeting capacity in road agencies. Local technical capacity can also be undercut by the frequent and sustained presence of foreign technical assistance that is funded under project budgets. Far from helping develop effective state bureau- cracies, certain aid practices reinforce the patrimonial elements within recipient governments. Projects provide all sorts of discretionary goods (such as vehicles, scholarships, international training, and so on) and are used to dispense favors by those in charge of the donor projects. The conclusion seems clear. Continued aid will be necessary to help Africa raise the performance of its transport infrastructure to interna- tional standards. But increased attention needs to be given to the creation of aid mechanisms and processes that will avoid the perverse effects of aid dependence. Human Resource Capacity At the time when most African countries gained their independence, in the late 1950s, only 16 percent of the adult population was literate. The entire region, with a population of 200 million, had produced only 8,000 secondary school graduates, over half of whom came from Ghana and Nigeria (Ajayi, Goma, and Johnson 1996). Yet decolonized Africa entered independence with an impressive array of higher education insti- tutions, set up by the colonial powers after World War II. The new gov- ernments saw these as primarily colonial remnants, irrelevant to their countries' immediate problems, and turned them into national universi- ties under strict political control. How to balance the allocation of scarce resources among different types of educational institutions has been a controversial topic for years. For more than two decades, the World Bank and other international financial institutions have focused on primary over higher education; some critics see this as a major cause of the decline of Africa's higher education sector (Samoff and Carrol 2004). 374 Africa's Transport Infrastructure By the late 1990s, the universities were in a state of crisis; it was argued that they could escape this crisis only through the restoration of academic freedom and institutional autonomy (Ajayi, Goma, and Johnson 1996). While the policies of the international financial institutions have moved back to favor higher education, and some improvements have occurred in recent years, the technical and managerial skills deficit is still substantial. More recently, attention has been focused on the inability of African countries to retain their high-skilled and tertiary-level-educated labor force (Marfouk 2008). Inadequate skills and leadership impose a serious constraint on policy development and operational efficiency in the transport sector. They also prevent the full use of existing technical capacity. Road authorities often lack the skills needed to review the design, costs, and work schedules of various contracts. This deficit prolongs the contracting process and could even be a reason for the recent escalation in the unit costs of road con- struction. At the heart of the problem is the poor payment of public ser- vants, which often encourages corruption and makes it difficult to retain critical skills in the public sector--as in air transport safety supervision. Overcoming this problem is not easy. For example, the road construc- tion industry in African countries has been dominated by large foreign- based firms, some of which operate in joint venture or in association with small local firms and a few medium-sized firms from the region. In recent years, the availability of staff from the downsizing of government depart- ments' in-house "force account" units has generated a large number of small domestic firms. But such firms have a low survival rate (Brushett and Seth 2005). Few small contracting firms have been able to grow to medium size because of (i) limited access to construction equipment, (ii) limited access to capital and credit facilities, and (iii) lack of busi- ness training and technical and management skills. Concerted effort is thus still needed to develop an indigenous contracting industry. The need has been recognized for some time. In 1993, a meeting of the Southern African Construction Industry Initiative reached consensus on the need to implement a national construction policy, to expand the role of domestic contractors and consultants through public-private partnerships, to study constraints on the development of the local road-construction industry, and to develop specific programs and measures to address these constraints. Subsequent regional initiatives through the Southern African Regional Construction Industry Council have proved ineffective relative to country-level initiatives such as the national construction councils estab- lished by Malawi and Zambia in the mid-1990s. That said, a national Governance: The Key to Progress 375 construction council can capture the ministry of construction, leading to fragmentation of contracts, collusion, and lower efficiency (as appears to be happening in Zambia). The usefulness of such entities thus depends on the existence of a suitable policy framework and a clear strategic vision and business plan. South Africa has the strongest program in the region, as con- struction industry development has become a critical element of the gov- ernment's strategy for economic empowerment of the majority. Training in road management and finance is also an urgent need for both the public and private sectors. With the assistance of the Sub- Saharan Africa Transport Policy Program, many senior executives have already received overseas training. But the involvement of regional asso- ciations is key to meeting a number of goals: strengthening program design; defining specific submarkets and developing relevant offerings to meet demand; promoting the wider involvement of training institu- tions, including those in Africa; and developing and disseminating mate- rials for communicating innovations and advice (Brushett, Sampson, and Waithaka 2004). Political instability also affects capacity--leaders whose hold on power is insecure seldom welcome the emergence of strong governing institu- tions that could serve as bases for the emergence of rival power. Such leaders would thus seek to undermine these institutions and put trusted individuals in control. In this context, the creation of road agencies has not always led to the expected improvement in sector performance, as agency leadership is frequently given to a politically connected candidate rather than one chosen competitively in an open international market. Other constraints on effective governance include the absence of timely and reliable data and information. Data on the transport sector are hard to come by, and when available are inconsistent, incomplete, or plain incorrect. Traditional styles of decision making thus too often rely on con- siderations of status, experience, or skill and leave significant room for subjectivity and discretion. Where internal control processes are also weak, this situation can be manipulated by corrupt agents to their own advantage. Information and communications technology thus has a big role to play in managing and processing data and providing access to information, and thus in making transport operations more transparent. The Way Forward Poor governance is seen to be a critical factor in the inadequate perform- ance of transport infrastructure in Africa. To a large extent, such governance 376 Africa's Transport Infrastructure reflects an economywide malaise that calls for reforms outside the trans- port sector. But there is also a range of sector-specific reforms in institu- tions, attitudes, and capacities now being implemented in some countries that may benefit the region as a whole. Priority 1. Institutional reform Given the failures of state enterprise and the dangers observed in relying on unregulated or self-regulated private sector supply, the suggested way forward is to concentrate on the establishment and strengthening of pub- lic sector planning and regulatory institutions to manage a predominantly private sector supply. While establishment of a general rule of law and some economy-wide antimonopoly powers may contribute to improved transport sector performance, most of these institutions will need to be sector specific. The most promising sources of sector improvements include the following: · Further strengthening of road fund institutions and procedures · Creation of commercially based highway implementation agencies · Creation of metropolitan transport authorities · Development of effective quality control procedures for trucking and urban bus systems · Extension of air license liberalization to domestic markets · Establishment of a direct financing mechanism for air traffic control services · Conversion of service ports into landlord ports · Creation of national transport safety councils. Priority 2. Changing attitudes Institutions are the organizational structures within which policy mak- ing and management take place. But if the prevailing attitudes of insti- tutional managers and other actors are inappropriate, even the best-structured institutions can fail. The most important issue to address is the reduction of discretion in various administrative proce- dures as a defense against corruption. The following appear to be the most significant changes in relevant attitudes or cultures suggested to improve sector performance: · Regular preparation and dissemination of reports (such as accounting and financial reports, asset inventories, annual reports) to aid institu- tional integrity Governance: The Key to Progress 377 · Streamlining of customs procedures to reduce discretion in adminis- tration · Development of formula-based procedures for the allocation of in- vestment and maintenance funds · Strengthening of concession award procedures · Development of new "negative concession" structures to facilitate pri- vate sector supply of loss-making services deemed socially necessary · Strengthening of supervision and monitoring mechanisms within the ministry of finance and parliament to enhance sector oversight. Priority 3. Improving implementation capacity The capacity to implement must also be improved. This priority funda- mentally requires adequate human resource capability to oversee how resources are allocated and employed. The following appear to be the most critical improvements required: · Reduced dependence on offline implementation units for aid projects · Development of national transport sector training programs · Establishment of national construction industry councils as nonexecu- tive agencies within a broader strategy for industry development · Review of public sector staff recruitment and remuneration arrange- ments to improve effectiveness and reduce staff turnover. Priority 4. Improving donor coordination Donor agencies are closely involved with governments in many countries in supporting capacity building and in preparing and approving invest- ment projects and programs. It is important that countries get the most that they can out of this involvement. The following steps are therefore suggested: · Governments should, alongside donors, regularly review the effective- ness of their capacity building. · Governments should encourage donors to accept common reporting systems, tender appraisal systems, and so on. Notes 1. This chapter benefited from input by Kavita Sethi and Bruce Thompson. 2. The indicators have not been without criticism. Arndt and Oman (2006) point out the problems of correlated errors, sample biases, and a lack of 378 Africa's Transport Infrastructure transparency, and they question the comparability of the indicators over time and across countries. While these indexes purportedly measure sepa- rate criteria, it has been argued that in practice they all tend to reflect a par- ticular perception of good governance (Langbein and Knack 2008). Their "construct validity"--whether they actually measure what they purport to measure--has also been questioned (Thomas 2010). Iqbal and Shah (2008) highlight some anomalous conclusions and argue that the indicators prima- rily capture Western business perspectives on governance processes and completely neglect citizens' evaluations, and that they should not be used in making cross-country comparisons. For these reasons, the discussion here focuses on the indicators' general pattern rather than on specific values. References Ajayi, J. F. A., L. K. H. Goma, and G. A. Johnson. 1996. The African Experience with Higher Education. London: James Currey. Arndt, C., and C. Oman. 2006. Uses and Abuses of Governance Indicators. Paris: Organisation for Economic Co-operation and Development. Brautigam, D. A., and S. Knack. 2004. "Foreign Aid, Institutions, and Governance in Sub-Saharan Africa." Economic Development and Cultural Change 52 (2): 255­86. Brushett, S., L. Sampson, and S. Waithaka. 2004. "Building Capacity in Management and Financing in the Road Sector: Meeting the Challenge." Sub- Saharan Africa Transport Policy Program Technical Note 37, World Bank, Washington, DC. Brushett, S., and S. Seth. 2005. "Construction Industry Development and the Road Sector Effectiveness of National Construction Councils." Sub-Saharan Africa Transport Policy Program Technical Note 38, World Bank, Washington, DC. Burnside, C., and D. Dollar. 1997. "Aid, Policies, and Growth." Policy Research Working Paper 569252, World Bank, Washington, DC. Gwilliam, K. M. 2007. "Paving the Road for Better Capacity: Review of Capacity Development in the Lao PDR Transport Sector." World Bank, Vientiane, Lao People's Democratic Republic. Iqbal, K., and A. Shah. 2008. "How Do Worldwide Governance Indicators Measure Up?" Paper presented at a World Bank seminar, Washington, DC, March 26. Kaufmann, D., A. Kraay, and M. Mastruzzi. 2010. "The Worldwide Governance Indicators: Methodology and Analytical Issues." Policy Research Working Paper 5430, World Bank, Washington, DC. Governance: The Key to Progress 379 Langbein, L., and S. Knack. 2008. "The Worldwide Governance Indicators and Tautology: Causally Related Separable Concepts, Indicators of a Common Cause, or Both?" Policy Research Working Paper Series 4669, World Bank, Washington, DC. Marfouk, L. 2008. "The African Brain Drain: Scope and Determinants." Working Paper DULBEA 08-07RS, Department of Applied Economics, Free University of Brussels. Meredith, M. 2005. The State of Africa: A History of Fifty Years of Independence. London: Free Press. Moss, T., and A. Subramanian. 2005. "After the Big Push? Fiscal and Institutional Implications of Large Aid Increases." Working Paper 71, Center for Global Development, Washington, DC. Nyerere, J. 1998. "Good Governance for Africa." Southern African Political and Economic Monthly. April. http://www.hartford-hwp.com/archives/30/083.html. Osborne, E. 2004. "Measuring Bad Governance." Cato Journal 23 (3): 403­22. Samoff, J., and B. Carrol. 2004. "Conditions, Coalitions, and Influence: The World Bank and Higher Education in Africa." Paper presented at the Annual Conference of the Comparative and International Education Society, Salt Lake City, March 8­12. Sequeira, S., and S. Djankov. 2008. "On the Waterfront: An Empirical Study of Corruption in Ports." Unpublished paper, Harvard University, Cambridge, MA. SSATP (Sub-Saharan Africa Transport Policy Program). 2007. "RMI Matrix for September 2007." World Bank, Washington, DC. Thomas, M. A. 2010. "What Do the Worldwide Governance Indicators Measure?" European Journal of Development Research 22: 31­54. Transparency International. 2009. Corruption Perceptions Index. Berlin: Transparency International. Transparency International, Kenya and Kenya National Commission on Human Rights. 2006. "Living Large: Counting the Cost of Official Extravagance in Kenya." Transparency International, Nairobi, Kenya. UNECA (United Nations Economic Commission for Africa). 2004. Accelerating Regional Integration in Africa. New York: UNECA. World Bank. 1989. Sub-Saharan Africa: From Crisis to Self-Sustainable Growth. Washington, DC: World Bank. ------. 2010a. Doing Business 2010. Washington, DC: World Bank. ------. 2010b. World Governance Indicators. Washington, DC: World Bank. CHAPTER 10 Conclusion: An Agenda for Action Africa has been shown to have less transport infrastructure per square kilo- meter than any other world region. Moreover, much of the infrastructure that it does have was designed to low standards (as in the case of railways) or is poorly equipped (ports and air traffic control). Yet because national incomes are so low, transport infrastructure expenditures constitute a rel- atively high proportion of gross domestic product (GDP) for many coun- tries. As a consequence, maintenance is often underfunded and the condition of infrastructure remains poorer than that in other regions. Africa needs better transport infrastructure, which will inevitably require more spending. Yet inadequate infrastructure is only one factor behind the poor performance of the transport sector. As the chapters in this book have shown, the quality of the service that transport infra- structure provides is critically dependent on the efficiency with which the infrastructure is maintained, managed, and used. Inefficient use of existing infrastructure raises expenditure needs and fiscal demands, while institutional and policy deficiencies inhibit the effectiveness of new investment. Improving transport quality thus depends not only on the level of expenditure but also on the appropriateness of the policies adopted and the quality of sector governance. This final chapter summarizes the main 381 382 Africa's Transport Infrastructure conclusions of the book on the issues relevant to policy, governance, and expenditure, with the first two of these areas presented as necessary con- ditions for properly identifying where infrastructure improvement is needed and for properly carrying out that improvement. Critical Transport Policy Issues This section sets the stage for a discussion of investment and fiscal needs by summarizing policy issues relevant to the more efficient use of trans- port infrastructure and services in the region. Some Multimodal Issues Some issues are multimodal because they apply either to all transport subsectors or to the relationships among sectors. Four such generic issues are discussed in the succeeding sections. Mobilizing competition effectively. There is ample evidence, particularly in the road and air transport sectors, that competition among modes of transport improves service quality and reduces costs. Such competition may also improve infrastructure efficiency through the replacement of one mode by another. For example, road improvement in Mauritania has effectively eliminated the domestic air transport sector, while all African railways have found it difficult to retain passenger traffic on an economi- cal basis when traveling by road is cheaper. The relationships among modes are rarely simple. Where modes com- pete, traffic should in theory be allocated among them on the basis of their relative price, in turn reflecting their relative costs. But in practice this is rarely the case. Where traffic is heavily imbalanced, as in inter- national trade, and cost structures differ across modes, both commer- cial pressure and economic efficiency may call for widely varying price-to-cost ratios. Taken to the extreme, however, such price discrim- ination may become predatory, giving one mode an advantage that is not justifiable from the point of view of the economy as a whole. Unfortunately, such predatory pricing is often difficult to distinguish from economically sensible price differentiation. This is complicated by the fact that modes may complement one another for some types of traffic while competing for others. Despite these conceptual problems, it is important to pay attention to competitive conditions. Roads, as public goods, are typically provided to users at costs that do not even cover maintenance, while privately Conclusion: An Agenda for Action 383 provided rail service is expected fully to cover its costs, including those of tracks. The viability of privately concessioned railways can be undermined by policies that undercharge road users and do not enforce truck load lim- its. National transport strategies that cover infrastructure costs of road and rail by charging users adequately for both would reduce states' budgetary burden while improving road conditions. The net effect of modal competition on the distribution of traffic across modes is often difficult to predict because of the importance of security, reliability, and other noncost items in determining transport choices. For example, a comparison of road tariffs and tariffs on five con- cessioned railways showed road tariffs exceeding rail tariffs by between 44 and 213 percent (Bullock 2009). Given the cartelization and high profit margins of road haulage, increasing road user charges might reduce truckers' profit margins without diverting any traffic from truck to rail. But it would nevertheless improve the sustainability of rail systems by increasing their pricing power in market segments where the railways have a real comparative advantage. By the same token, substantial bene- fits could likely be obtained by promoting competition within the road haulage sector. Revisiting attitudes toward private supply and profit. Private participation in supply is central to competition and can increase the efficiency of oper- ations and the mobilization of private capital. But to tap the potential of private participation, governments need to understand the commercial real- ities that motivate private business. In the interest of consumers, monopo- listic behavior should be constrained. But private firms will not participate if governments deny them a reasonable return on their investment-- making a profit is not a crime; rather, it is a necessity for doing business and a proper incentive to allocate resources efficiently. Plans to attract private sector finance and management should include an explicit determination of the objectives of private participation and the reasons for seeking it. Policy makers should recognize that private par- ticipation can bring efficiency benefits not only to transport modes with marginal commercial viability but also to those that are highly profitable (ports) or deemed desirable but highly unprofitable (some railway con- cessions). To attract private capital, policy makers should consider a wider range of scenarios for participation, including negative concessions and affermage arrangements. A national agency focused on privatization and its promotion could help in producing such policies and outlining areas of participation. 384 Africa's Transport Infrastructure Public monopoly powers are exploited to generate government rev- enues from ports, airports, and air transport services in several countries. But it can be damaging for the government to exploit monopoly powers to generate excess revenues, even to support other unprofitable services. Internal cross-subsidies usually have adverse impacts. For example, South Africa's Transnet Freight Rail is organized to draw cross-subsidies for loss-making freight and passenger services both from the profitable ore and coal services and, through the Transnet group, from ports and pipelines. But the rail's core services and infrastructure have been weak- ened as a result. Locomotives are on average 25 years old and freight wagons are 25 to 30 years old--ages nearly double those recommended by international best practice. There is a capacity gap in the ore services, and the safety record of Transnet Freight Rail is low and appears to be deteriorating. The port system is also being denied investment. Despite the high technical competence of the South African system, the recent National Freight Logistics Study concluded, "restoring rail reliability is fundamental and is the single most important challenge facing the freight logistics sector in South Africa" (South African Department of Transport 2007). Careful consideration thus needs to be given to establishing appropri- ate oversight and regulatory institutions. The creation and exploitation of monopoly powers by cartelization, particularly in the trucking sector, needs to be continually reviewed, with assessments covering both indus- trial structure and commercial behavior. Given the scope of this task, many countries could develop a small but skilled regulatory unit to advise gov- ernments, as well as regulators specific to particular modes. One issue related to foreign private sector involvement in supply mer- its consideration. Concerns about reliance on foreign control of services critical to national security have been widely used as an argument for maintaining national air transport and shipping fleets and for limiting for- eign capital in national infrastructure finance. In practice, uneconomical small airlines and shipping fleets protected by cargo reservations tend to push up costs and drain national resources (to the detriment of national security). And resistance to involving the global container-port terminal operators denies countries the efficiency and investment that such partic- ipation could bring. Countries thus need mechanisms to reconcile private (and foreign) financing of transport infrastructure with economic, social, and national strategic objectives. Developing a range of contract designs appropriate to a range of objectives is an important part of this challenge. Conclusion: An Agenda for Action 385 Developing modern logistics systems. Logistics systems in Africa are viewed as poor by business users. In 2007, the Logistics Performance Index showed only South Africa in the top quartile of countries and only 5 of the 39 African countries in the top half; the situation in 2010 is vir- tually unchanged (World Bank 2007, 2010).1 Africa would be the worst- performing region in the world if Afghanistan did not drag down South Asia's unweighted average of only six countries. The two dimensions in which Africa scores the lowest are infrastructure and customs. Physical infrastructure needs are discussed below. As far as customs are concerned, the problem is largely an issue of corruption. In the Corrup- tion Perceptions Index (Transparency International 2009), only Botswana appears in the top quartile of countries, while 17 African countries appear in the bottom quartile (more than one-third of the countries in that quar- tile). These statistics highlight the need to come to terms with the prob- lems at border customs posts in the region. A shift toward regional free trade would help. But even without that, the successes achieved in south- east Europe--both through a comprehensive attack on the problems involving physical infrastructure, staffing, administrative procedures, and so on, and through regionwide agreements among countries--show that the problems are not insuperable. Other factors in addition to infrastructure erode logistics system qual- ity. Restrictions on the freedom to select haulers undermine the ability of third-party logistics service suppliers to operate a one-stop shop for inter- national freight movements, of crucial importance to modern logistics systems. Weakness in information technology systems also makes moni- toring and control of freight movements less effective. The lack of these supporting services severely impedes the growth of a third-party logistics sector, which could contribute to the development of global manufactur- ing and distribution chains. Taking transport safety seriously. Inadequate safety provision is a prob- lem across all modes, and the region has arguably the worst safety record in the world in the road, rail, and air transport sectors. The general lack of a culture of transport safety can be attributed to two factors common to all modes. First, necessary safety regulatory institutions are weak or non- existent. Second, in the absence of adequate supervision, operator behav- ior is frequently dangerous. Other causes differ somewhat by mode. Rail accidents frequently result from the poor state of infrastructure. Road accidents are more typically associated with the poor separation of traffic movement from roadside activity and poor driving behavior compounded 386 Africa's Transport Infrastructure by lax enforcement. Air transport accidents appear to be associated more with poor crew training and lax supervision than with a lack of navigation aids or new aircraft. Whatever the sector, however, there is evidence to indicate that the problems are not insuperable if they are given adequate attention and priority. The various sectorwide policy requirements, and the agenda for action that they suggest, are summarized in table 10.1. Mode-Specific Policy Issues In addition to the common issues discussed above, the various transport modes suffer from a range of modally specific problems. Roads and road transport. The critical policy issues surrounding roads and the road transport sector concern financing maintenance, prioritizing Table 10.1 Sectorwide Policy Requirements: An Action Program Institutional Policy Topic Issue requirement requirement Competitive Unfair competition Give a single agency Subject issues of conditions among modes (a single ministry of modal interaction transport or national to explicit policy transport council) review. responsibility for issues of modal coordination. Attitudes toward Unrealistic Establish competitively Focus government private sector expectations of tendered concessions, supervision on private sector and treat all obligations monitoring and behavior and payments as enforcing contracts enforceable contractual rather than on conditions. day-to-day intervention. Logistics quality Poor logistics Create a national Reform customs systems logistics council to administration. development advise on trade Liberalize entry to requirements. road haulage markets. Liberalize telecom- munications markets. Transport safety Absence of a Create a national Make safety an issue general safety transport safety council at the highest level. culture and subsector-specific Develop subsector agencies. safety programs. Source: Authors. Conclusion: An Agenda for Action 387 expenditures, increasing the efficiency of roadwork implementation, improving safety, and regulating haulage operations. Road maintenance financing has been greatly improved in many coun- tries by the establishment of road funds. But as shown in chapters 2 and 8, many countries still do not provide sufficient funding to maintain the road system in its current state of repair, much less to improve it and eliminate backlogs. While the road fund board may sometimes underesti- mate how much finance is needed, it is more likely that the board is con- strained by the finance ministry, either formally or informally, to set the fuel levy at a level insufficient to meet maintenance needs. Prioritizing road expenditures involves maintenance as well as invest- ment decisions, and depends on complex judgments. In countries where the value of road assets is high in relation to GDP (as in Malawi and Namibia, discussed in chapter 2), the current road network may simply be too extensive to maintain. In these circumstances, rather than letting the whole network deteriorate, it may be better to deliberately abandon part of the network (perhaps encouraging communities to maintain it) to save the rest. Even in cases where such a drastic step is not necessary, it is sen- sible to relate the character and condition of the roads to traffic levels. Chapter 2 has shown that in most countries the most highly trafficked roads are designed to higher standards and maintained in better condition than the more lightly used roads. But the Road Network Evaluation Tool (RONET) analyses also show some evidence of overengineering on the main road networks and underengineering on the rural networks, suggest- ing the need for a change in priorities. Where to devote resources depends on the relative weights given to ensuring rural accessibility versus maxi- mizing the short-term economic benefit of the (interurban) road system. Roadwork implementation has been shown to be more efficient when done by a private sector contractor than when countries rely on force accounts. In turn, results-based maintenance contracts have been shown to be an effective way of mobilizing the private sector. There is now a move in some countries toward road asset management contracts that are even more comprehensive. Such countries must be able to enter into long-term contracts (facilitated by a road fund) and be unconstrained by a commit- ment to force accounts (facilitated by a quasi-autonomous roads agency). The establishment of such institutional arrangements is advisable. Road safety, though a perennial problem, has only recently been rec- ognized at the highest level of intergovernmental deliberations. At the 2007 Pan-African Road Safety Conference, participants resolved to set road safety as a national health and transport priority and elaborated a 388 Africa's Transport Infrastructure wide range of possible policy instruments to promote safety. But for these to be put into effect, most governments still need to establish an appro- priate institutional framework (as attempted in Ghana) and to generate strong local impetus for a comprehensive road safety program (as attempted in KwaZulu-Natal, South Africa). Cartelization and protection in the road freight sector have been shown to increase profits and prices, particularly in Central and West Africa. Governments condone, indeed encourage, such behavior by col- laborating with truckers' associations in the administered, noncompeti- tive allocation of transit freight from ports, on the pretense that this is necessary to ensure a fair allocation of traffic between "home" and foreign carriers. These problems would be solved if entry into the industry were restricted only by quality licensing, and cartelization were subject to national antimonopoly legislation. The main issues requiring action in the roads and road transport sector are summarized in table 10.2. Table 10.2 Roads and Road Transport: An Action Program Institutional Policy Topic Issue requirement requirement Maintenance Insufficient Ensure a private Have road boards finance fuel levy sector majority produce annual on road boards. estimate of finance requirement. Prioritization Over- or Establish a national Have road boards of road underengineering "project preparation develop explicit expenditures pool" subject to a policy on fund common appraisal allocation. process. Road work High cost Establish Increase private implementation Poor quality quasi-independent contracting. road agency. Move to performance- Abolish force accounts. based maintenance contracting. Road safety Very high Establish a national Have NRSC launch accident rates road safety council intensive road-safety (NRSC). campaign. Road freight Very high road Restrict entry only by Abandon transport freight rates quality licensing. administered regulation Make road haulage allocation of transit subject to import traffics. competition law. Source: Authors. Conclusion: An Agenda for Action 389 Rail transport. Most African railways were developed by colonial admin- istrations to assist in the exploitation and export of agricultural and min- eral resources. The most viable lines are still those specializing in the export of minerals from South Africa, Gabon, and to a lesser extent Zambia. With the exception of the South African (and more recently the Nigerian) urban commuter services, railways play only a small role in the passenger transport market. Given their limited social role, they have been seen as particularly suitable for concessioning to the private sector. Relevant policy issues mainly concern the suitability of the forms of concession adopted. Public service obligations are frequently imposed on railways, whether they are state owned or concessioned to the private sector. In some cases, such obligations are imposed even though there are cheaper road services available. Some countries have agreed to pay rail operators to compensate for the imposition. But as discussed in chapter 3, these payments are rarely made in an adequate and timely manner. The pol- icy response to this situation needs to involve not only a formal analy- sis of whether public rail service is a necessity but also a legal obligation for the state to compensate the private sector on a contractually pre- determined basis. Government capital contributions to concessions may be a way for- ward. As argued earlier, low-interest loans from international financial institutions to railway concessionaires have tended to disguise the real financial burdens of system maintenance in the long term. Chapter 3 showed that only lines with a density of 2 to 3 million net tonnes or more can fund full rehabilitation from a purely commercial viewpoint (including the costs of capital). But if governments were to bear the costs of capital, that break-even traffic volume would fall to below 1 million net tonnes per year. Hence, if governments wish to retain a public railway service, they will need to find some way of sharing the capital cost burden and set up contractual terms different from those employed to date. Competition from the road sector has a significant effect on rail con- cession finances. Two linked elements of that competition are of particu- lar concern. First, failure to enforce axle-load limits enables truck rates to compete "unfairly" with the railways. Second, overloading trucks increases road wear and maintenance costs. In many countries, the heavy trucks that cause such damage are not appropriately charged for it. The main issues requiring action in the rail transport sector are sum- marized in table 10.3. 390 Africa's Transport Infrastructure Table 10.3 Rail Transport: An Action Program Institutional Policy Topic Issue requirement requirement Public service Unrealistic Include any public Pay contract obligations obligations expectations of service obligations on time. concessionaires' in contracts, with ability to compensation cross-subsidize arrangements clearly specified. Investment Inadequate Develop "negative" Clearly specify long-term finance provision for concession or investment requirements investment affermage models, in invitations to tender. with government participation in investment finance. Competitive Unequal Ensure that Have governments environment treatment of institutions or participate in finance of track costs for procedures for infrastructure investment road and rail examining the and/or increase charges modes consistency of for road use. modal policies exist at the central government level. Source: Authors. Airports and air transport. Air transport is less well developed in Africa than in other regions and has typically been provided primarily by the flag carriers of the former colonial powers. Though new carriers are enter- ing the market, and three major African airlines have a substantial share of it, some serious problems remain. The critical issues for the air trans- port sector relate to airport investment strategy, air traffic control (ATC), air service provision, and air transport safety. Airport investment strategy requires careful reconsideration. Minor investments in taxiways and strategic rescheduling of flights to spread air- port movements over a broader time period can in many cases overcome what are perceived as runway capacity limits. Such actions would free up resources for the improvement of landside facilities (which would also relieve the pressure at peak periods). If airlines were charged a modest premium for landing at peak periods, they would be more likely to alter their schedules. Air traffic control services were shown in chapter 4 to be seriously deficient. This is because civil aviation authorities (CAAs) are expected to support themselves through fees, yet in many cases a percentage of Conclusion: An Agenda for Action 391 these fees goes to the treasury. Not only are such authorities short of cash, but they find it very difficult to retain highly trained staff on public sector salaries. There are three possible steps toward solving this problem. First, it is likely that costs could be significantly reduced by replacing radar installations with more advanced, satellite-based technologies--careful attention needs to be given to what would be gained and lost by such a shift. Second, the development of commercial regional pooling arrange- ments for air traffic control might both reduce costs and better retain staff. And third, the CAAs could be guaranteed a more secure, predetermined share of revenue from air traffic control. Air transport service regulation also needs reform. National flag carri- ers, historically protected by government negotiations of bilateral landing rights in the international and intercontinental markets, are no longer able to cross-subsidize domestic markets and should not be protected per se. Indeed, many of these carriers have failed, and air connectivity has suf- fered as a result. A new strategy needs to be based on extending the lib- eralization that has already occurred in the international market in Africa to all other sectors, together with developing hub-and-spoke networks based on major regional (not just national) airports. Employment of small turboprop aircraft could improve the economic viability of small airlines and thin routes, but only within an appropriate economic context. Air transport in Africa has a very poor safety record, mostly confined to indigenous African air operators. The record of the major international operators is not significantly worse in Africa than elsewhere. While the average age of domestic operators' fleets is high compared with fleets in other regions, and there is a high proportion of old Eastern-bloc aircraft, the real source of the problem, according to a number of independent sources, is inadequate training and supervision. Skilled staff are difficult to hire and retain, and the supervision of both flight crew and maintenance staff appears to be below international standards. To some extent, these problems may be helped by greater liberalization, which would generate a more economically viable sector. But liberalization needs to be accom- panied by much stricter enforcement of safety standards. To achieve this, civil aviation authorities will need to be strengthened and critical staff will need to be paid more. The main issues requiring action in the air transport sector are summa- rized in table 10.4. Ports and maritime transport. There is some disagreement over the sta- tus of current port capacity, with Drewry Shipping Consultants (2009) in 392 Africa's Transport Infrastructure Table 10.4 Airports and Air Transport: An Action Program Institutional Policy Topic Issue requirement requirement Airport Selection of Strengthen Look for investment investment economic skills improvements strategy program of CAAs. in existing airports rather than entirely new locations. ATC and Adequacy of Include provision Recognize need for navigation payments for payments for a secure funding services ATC and air source for ATC navigation services and ANS. (ANS) in the CAA law. Air transport Failure of national Close small Liberalize entry to service flag carriers loss-making domestic markets. regulation Loss of service on less national flag Subsidize unprofitable profitable routes carriers. routes through Poor network of competitive tenders. regional services Encourage hub-and-spoke services through major regional airports. Air transport Poor safety record Strengthen CAAs. Increase commitment safety of some indigenous to enforcement of African carriers safety legislation. Source: Authors. particular arguing that quay capacity is for the most part adequate. The question turns on the efficiency with which goods are handled on quay. African ports in general have handling rates--both of containers and of dry bulk traffic--considerably below world standards. Investment in quay space and improvement in handling efficiency are not mutually exclusive, but one does not necessarily follow on the other. The question is how to get the best combination. Improved port management can be obtained in many ports by moving from the traditional public service port model to the landlord model. The concessioning of facilities, in particular container terminals, has been shown in chapter 5 to be a common source of improvement. But conces- sioning is not always easy, and there have been several cases of litigation relating to poor concessioning procedures. Governments should therefore ensure that these procedures are clearly defined and transparent. Extended use of the major international container-terminal management companies or port managers is recommended. The development of port Conclusion: An Agenda for Action 393 community systems has been shown to lead to better understanding, communication, and data exchange, and it is also important for security.2 Customs and immigration reform can often contribute to the effective capacity of ports by reducing the standing time of goods in port and thus increasing warehouse or terminal standing-area throughputs. Coordination with land transport is also poor in many ports. This is in part because the responsibility for coordination does not fall to any single party. Concessioning ports and railways to the same agency may help with coordination but risks creating excessive monopoly power. The main issues requiring action in the ports and maritime transport sector are summarized in table 10.5. Urban transport. Urban transport Africa is frequently chaotic. The urban road system suffers from inadequate funding (it usually gets only a small share of national road fund revenues), poor traffic discipline, high acci- dent levels, and public transport provided primarily by an informal sector subject only to self-regulation, with adverse effects on users. Above all, it Table 10.5 Ports and Maritime Transport: An Action Program Institutional Policy Topic Issue requirement requirement Port Poor port Engage in port reform Develop new, management performance and adopt the transparent landlord port concessioning management procedures. model. Concession major terminals. Customs High costs and Adopt customs Increase use of administration long delays systems and information in customs procedures of technology and clearance the World Customs electronic data Organization and processing to United Nations reduce corruption. Conference on Trade and Development. Coordination No clear location Establish a port Encourage effective with inland of responsibility community system. coordination transport for modal through appropriate coordination regulatory and (access roads pricing measures. and so on) Source: Authors. 394 Africa's Transport Infrastructure lacks any strategic vision or planning framework and requires a number of policy initiatives. Low traffic speed and high accident rates have a number of roots. Road space is invaded by traders and pedestrians. Traffic composition tends to be very mixed. And there is usually very little attention given to enforcement of traffic discipline. Overcoming these problems requires appropriate institutions--particularly a strong strategic authority and a technically competent traffic management unit. But it also requires the involvement of the police and a wholehearted commitment to enforce- ment of traffic rules. The regulation of passenger transport fares has in many countries had the perverse effects of bankrupting the large bus companies that have provided services in the past and of giving rise to informal sector opera- tions that often charge higher rates. Attempts to rescue or restore the for- mal sector operators have usually failed because governments lack the funds to support them. What is needed instead is a competitive process for procuring bus services, one that makes the costs transparent from the outset. In most countries, this would require substantial institutional and policy reform. Self-regulation of minibus transport is usually motivated by the desire to avoid predatory on-the-road behavior, such as racing and blocking, and to ensure an equitable distribution of income among members of operators' associations. But while it can reduce dangerous driving behavior, it usually does so through the imposition of tour de role dispatching procedures, which reduce vehicle utilization and hence increase fares. The introduction of tendering processes to eliminate both predatory road behavior and inefficient dispatching procedures could solve this problem. The main issues requiring action in the urban transport sector are sum- marized in table 10.6. Improving Governance Many of the inefficiencies in transport infrastructure arise not because of failures to recognize appropriate policies but because of failures to implement them--in other words, failures of governance. Sometimes these failures occur because there is no appropriate instrument to implement policy (institutional failure) and sometimes because gov- ernment serves its own narrow interests rather than the interests of the governed. Conclusion: An Agenda for Action 395 Table 10.6 Urban Transport: An Action Program Institutional Policy Topic Issue requirement requirement Road Increased Create a strong Allocate adequate maintenance operating costs municipal or funding either from and greater metropolitan road fund allocations risk of accidents transport agency. or from local or arising from badly national tax sources. maintained roads Bus fare Decline of many Establish an urban Use competitive regulation companies public transport tendering of services because of authority to plan along with any fare unrealistically and procure services. controls. low fares Minibus Undesirable Establish a metropolitan Monitor behavior and service operating or municipal regulator enforce regulation regulation practices of to oversee minibus effectively. private operations. associations Road traffic High congestion Establish a traffic Deal with invasion of management and road management unit road space and accident levels within the municipal enforce traffic rules. or metropolitan transport agency. Source: Authors. Institutional Weaknesses Four main institutional weaknesses can be identified: the absence of any sectorwide strategic planning body, ineffective regulation of transport service providers, excessive aid dependency, and inadequate implementa- tion capacity. The strategic planning void. The inconsistency among subsector policies-- particularly between road and rail infrastructure charging policies--has been noted earlier. The problem is that different modes are often han- dled by separate government ministries or by departments within a sin- gle ministry that are not well coordinated. One solution would be to establish either a national transport council (to coordinate the actions of several ministries) or a national transport policy committee (to over- come excessive departmental autonomy and encourage coordination within a single central ministry). In either case, the agency must be estab- lished at a high level. 396 Africa's Transport Infrastructure The problem is particularly severe in urban areas, where the modes are at their most interdependent. Strategic metropolitan transport authorities are very rare, and where they do exist they may be depend- ent on other authorities for implementation. Although such agencies have been established in recent years in Dakar, Lagos, and other cities, they are typically still weak in both power and competence. The policy challenge is to establish a metropolitan authority with sufficient funds to perform its planning role effectively and sufficient leverage over the implementing authorities to ensure that their actions are in line with the metropolitan strategy. Ineffective regulation. Effective regulation entails more than requiring that private service suppliers assume and carry out social obligations; regulators must devise mechanisms for getting the most out of private operators--within the constraints of resource availability. To do so requires understanding how commercial transport suppliers think and how to design regulatory systems that reconcile the search for profit with the achievement of socially desirable service structures. Competitively tendering franchises or concessions is a suitable way to achieve that reconciliation. Faced by severe budget constraints, many African governments have turned to the private sector for both capital finance and management skills. This shift has brought its own problems. Where subsector control is highly centralized, as is common for railways, there is danger of developing a monopoly. Where the market structure is naturally frag- mented, as in the trucking or urban minibus systems, there is danger of promoting predatory competitive practices. In both cases, the tempta- tion is to arbitrarily impose fare or service requirements on private companies without regard to the realities of private business finance. Such heavy-handed regulation has often had very damaging effects on the quantity of service provided. Aid dependence. The analysis of the transport sector's revenue sources provided in chapter 8 showed a high level of aid dependence in several countries. In the long term, the reduction of aid dependence will be both a consequence and a symbol of the growth of national economies and the maturity of transport sector policies. In the interim, aid dependence can have a number of adverse effects, some of which were set out in chapter 9. The tendency of international financial institutions to finance rehabilitation but not routine or periodic maintenance amounts to a Conclusion: An Agenda for Action 397 disincentive to maintain. Cheap capital lent to finance rail concessions can disguise the long-term implications of concession finance. And in mounting large rehabilitation programs, both foreign technical assis- tance and highly skilled indigenous workers are too frequently absorbed in special project implementation units, a result that limits the dissem- ination of vital skills. All of these tendencies need to be resisted as part of a policy to avoid the possible adverse consequences of aid dependence. There are signs of change. The World Bank's development policy loans, which are attached to the development of policy-based programs rather than specific investments, have the potential to support maintenance- oriented reforms.3 Similarly, the European Union is experimenting with sector budget support in the road sector in Ethiopia, Madagascar, and Tanzania in which finance covers all activities, including maintenance, within a medium-term expenditure framework. In general, govern- ments and aid agencies need to work together to ensure that aid does not inhibit the development of good, sustainable national policies and practices. Implementation capacity. Finally, improving capacity at all levels will require a stronger formal education system and targeted improvements in both training and retraining sector workers with key technical and mana- gerial skills. In several areas, such as air traffic safety inspections, retaining highly skilled labor will depend on making public sector salaries compa- rable to those in the private sector. The main issues requiring action to overcome institutional weaknesses are summarized in table 10.7. Behavioral Failures In many cases, it is not the laws or even the institutions themselves that are inherently defective but the way in which the laws are interpreted and the institutions managed. A number of particularly egregious exam- ples of behavioral failure exist in the transport sector. Corruption. Corruption is rife. At the operational level, poorly paid offi- cials create artificial delays in customs administration to extract bribes from operators for speedier service. This practice increases transport costs and adversely affects competitiveness. But it is very difficult to contain this petty corruption when much more elaborate corrupt practices remain at the higher levels of government. Decades of dictatorial rule 398 Table 10.7 Overcoming Institutional Weaknesses: An Action Program Institutional Policy Topic Issue requirement requirement Strategic planning Absence of mechanisms for the Create a committee or council for Commit adequate resources coordination of mode-specific coordinating central transport policy. and powers for the strategic policies at urban and national Create a metropolitan transport agency. planning body at the national levels or metropolitan level. Regulation Perverse effects of ill-informed or Separate regulatory and service supply Set general government objectives heavy-handed economic regulation functions. and avoid political intervention Establish professionally staffed by government in day-to-day procurement and regulatory agencies. administration. Aid dependence Institutionalization of perverse Establish strong institutions for sector Establish formal resource allocation incentives regulation and administration. procedures. Incorporate official development assistance in line-agency administration rather than in institutionally separate project management units (PMUs). Implementation capacity Weak skills and education Strengthen higher and technical Recruit based on merit. Loss of skilled staff education. Avoid differential payment levels Failure to achieve effective skills transfer Review salaries for high-skilled for PMU staff. occupations. Inject all technical support into Phase out off-line PMUs. line management. Source: Authors. Conclusion: An Agenda for Action 399 buttressed by complex institutionalized systems of patronage have in many countries destroyed the integrity of public administration and the management capability of public enterprises. Even corporate private enter- prise is driven by patronage and special interests rather than the goal of productive efficiency. Corruption is essentially a behavioral issue: address- ing it has as much to do with applying existing anticorruption laws as it does with creating new ones. Despite the unfavorable environment in which transport infrastructure institutions work, there is already some evidence that actions at the sec- tor level can lead to improved behavior and performance. · First, attempts can be made to eradicate situations that combine mo- nopoly and administrative discretion--the institutional circumstances in which corruption thrives. Computerization of customs arrange- ments, vehicle inspections, driver's licenses, and other official docu- ments can have this effect, and has been successful in reducing corruption in developing countries elsewhere. ·Second, liberalizing transport markets replaces incentives that protect corrupt operators with incentives that increase efficiency. Liberaliza- tion has already brought about significant improvements in the shipping and international air transport markets, but there remains considerable scope in domestic air transport and, in some countries, in trucking. · Third, transparent management and regulatory institutions, working according to specified rules and subject to regular audits, tend to im- prove the efficiency of resource use, as exemplified by the develop- ment of road funding in road maintenance. Failure to finance maintenance. Road networks throughout the region are inadequately maintained. Yet much higher rates of economic return can be obtained on maintenance investment than on most investment in new roads. Lack of maintenance is sometimes defended on the grounds that the internal rate of return to the road agency budget of increasing periodic maintenance expenditures may not exceed the rates of return in other sectors--especially when cheap funds can be obtained from the international financial institutions for rehabilitation (but not for mainte- nance). But this claim does not hold true from the viewpoint of the national economy, as explained in chapter 8, because the benefits of timely maintenance accrue primarily to road users and not to the agency itself. Failure to recognize the importance of maintenance is a major prob- lem presently afflicting the sector. 400 Africa's Transport Infrastructure Failure to compensate for the meeting of social obligations. Most African governments, wishing to keep public passenger transport costs down, have constrained both bus and rail passenger fares. But when fares are set at commercially unviable levels, they are likely to undermine the supply of public passenger services. Even when the rates are enforced only on publicly provided modes (notably rail or conventional large-bus services), they typically result in a shift of business to an informal sector with higher fares or lower service quality, thus failing to achieve their stated objective of assisting poorer citizens. The only way of avoiding this outcome is for governments to compensate transport suppliers for the social obligations that they impose on them. Some governments already accept the need for such compensation. But only in rare cases are payments made at an adequate level and in a timely manner. Such payments are made in Addis Ababa, where the gov- ernment successfully supports public bus service, but elsewhere sporadi- cally paid subsidies merely postpone the eventual collapse of public bus services. The same appears to be true with the subsidization of rail pas- sengers in some of the rail concessions. Where subsidized rail services have economically viable road alternatives, the subsidy drains resources from the economy. More-comprehensive and well-thought-out strategies for fare controls and subsidies are necessary to ensure the most effective use of transport infrastructure. The procurement of subsidized services by competitive tendering is one likely solution. The absence of a safety culture. It is worth reiterating here that the scourge of transport accidents across the region can and should be the tar- get of institutional action. Yet only changes in the behavior of individual transport users promise to significantly improve safety. The challenges of reeducation and safety enforcement may be among the most intractable facing the transport sector in Africa. The main issues requiring action to overcome behavioral weaknesses in the transport sector are summarized in table 10.8. Expenditure Requirements The financial requirements of maintaining and improving transport infrastructure have been explored in two parts. First, in chapter 7, some estimates were made of the total expenditures necessary if the region is to achieve reasonable levels of infrastructure quality. Second, in chapter 8, these estimates were compared with recent expenditure Conclusion: An Agenda for Action 401 Table 10.8 Overcoming Behavioral Weaknesses: An Action Program Institutional Policy Topic Issue requirement requirement Corruption Corrupt practice at Establish anticorruption Facilitate greater all levels of public institutions at a freedom of entry administration national level. into the transport Establish independent business. auditing systems for Formalize rules public administrations. and computerize administration processes. Maintenance Underprovision of Strengthen road fund Commit to full finance finance for asset administrations. social cost-benefit maintenance appraisal of maintenance programs. Public service Adverse effects of Enable competitive Commit to making obligation (PSO) social service tendering of bus agreed-upon compensation requirements on and air social service payments on time. system viability contracts. Establish formal procedures for review of PSO compensation in rail concessions. Transport safety High accident Establish a national Develop and fatality rates, safety council. comprehensive particularly on programs in roads transport safety behavior. Source: Authors. levels to demonstrate the size and nature of the funding gap, and possible financing sources were discussed. Estimating Financing Needs Transport financing needs have been looked at in two different ways in this book. First, the mode-specific chapters discussed some of the main expenditure needs by subsector, though generally without attempting to identify and cost an optimum program. Those chapters also identified a number of investments that might seem politically attractive but are not likely to yield very good value for the money. In this category of potential white elephants were major new airports (when improving existing ones would be more economical) as well as some elements of the idealistic 402 Africa's Transport Infrastructure Pan-African rail or road networks. But it was not possible within the scope of this book to undertake the systematic project-by-project economic evaluation that would be necessary to construct a set of prioritized invest- ment projects. What was possible, however, was to see the obviously high economic returns promised by expenditures on improved maintenance, not only for roads but also for other modes. Given the impossibility of providing a complete economic evaluation, chapter 7 estimated the needs for transport infrastructure expenditures on a different basis. Based initially on an examination of networks in more highly developed countries, a set of "connectivity targets" was postulated, covering regional, national, urban, and rural connectivity and including all the major modes. Networks based on these standards were compared with existing infrastructure networks to identify, by subtraction, a menu of expenditure needs. In some cases, this involved upgrading existing net- work capacity or conditions, while in others, it involved investing in new facilities. Observed costs of infrastructure works in Africa were applied to convert the program into a financial bill, and a 10-year program period was postulated for the achievement of the selected standards. The "base" scenario specified this way was estimated to cost the region a total of $234 billion over 10 years, which would amount to 3.6 percent of GDP per year. For 13 countries, nearly all low income, the basic requirement would exceed 5.0 percent of yearly GDP over the period, with an average of 10.1 percent. Even including the relatively high levels of official development assistance (ODA) in the financing of transport infrastructure in low-income countries in recent years, this appears to be an unachievable objective. For this reason, a more modest set of standards was postulated--the "pragmatic" scenario. In this scenario, the investment needed for the 13 countries with the highest requirements was reduced from 10.1 percent of GDP to 6.1 percent on average. This regional target seems more achievable, and so it was treated as the basic need in the subsequent analysis. In total, in both the basic and the pragmatic scenarios, about 90 percent of the estimated expenditure needs are in the road sector, with that total fairly evenly divided among regional, national, local access, and urban roads (with the urban share falling a little in the pragmatic scenario). Most significant, the expenditure needs assess- ment concluded that 40 to 50 percent of the expenditure required should be spent on maintenance--a proportion that the region over- all is not even close to achieving. Conclusion: An Agenda for Action 403 Securing Finance The detailed analysis of the revenues and expenditures of 24 countries, described in chapter 8, showed average total annual expenditures on trans- port infrastructure in recent years contributing $16.2 billion toward the estimated needs (in the scaled-down pragmatic scenario) of $18.2 billion. This is 2.5 percent of national GDP on average. At first glance, it appears that a surprisingly high proportion of needs are being met. The analysis also identifies three important inefficiencies in the financing arrangements, which, if addressed, could reduce the funding deficit even further. These are the undercollection of fees from users, the underexecution of allocated investment funding, and the underallo- cation of resources to routine and periodic maintenance. Addressing the first two of these issues would directly improve the finances of the implementing agencies. The third is rather different, as the effect of undermaintenance is to increase vehicle operating costs, so the benefits of addressing it would accrue to road users rather than to the imple- menting agencies. A number of caveats are needed to qualify this apparently promis- ing scenario: · First, considerable uncertainty attaches to some unit costs, particularly of rehabilitation, partly because reliable data are difficult to obtain (see chapter 7). · Second, the picture looks far less favorable if ODA is excluded from the calculation. This accounts for nearly 25 percent of the capital expenditure and about 12 percent of the total expenditure. · Third, the picture varies enormously both among country groups and among individual countries. The deficits for many low-income countries and all low-income fragile states are much greater than the average, and the situation in many of these countries remains dire. Reallocation of funds among countries is likely to be an option only if a reallocation of ODA support is pursued. · Fourth, the benefits of addressing the underallocation to periodic and routine maintenance, which would accrue primarily to infrastructure users, could improve road agency finances only if accompanied by some tax or user fee reforms. · Fifth, the targets have already been scaled down overall in the "prag- matic" scenario, so the balance between finance and needs may seem unduly favorable for higher-income countries, which have higher cur- rent standards and aspirations. 404 Africa's Transport Infrastructure These caveats emphasize that many countries will need to pursue additional ways of raising funds for transport infrastructure. The imme- diate prospects for finding these other sources are difficult to assess. Government is the main source of funding in most countries, but given the wide range of claims on the budget, the likelihood of significant increases in funding from this source seems low. Most funding has come from mem- bers of the Organisation for Economic Co-operation and Development. Other funding has come from Chinese sources associated with natural resource exploitation, from India, and from the Gulf states. Funding from sources other than the Organisation for Economic Co-operation and Development has been focused on the rail sector, and its robustness remains to be seen. Private capital flows have been volatile over time and have been focused mainly on middle-income, resource-rich countries, where they have gone to seaports and airports, as well as to rail conces- sions and a few toll roads. These flows are under stress during the global economic downturn. Local capital markets remain relatively weak, and the available maturities are often too short to be attractive for transport infrastructure investment. That leaves ODA, which was already scaled up substantially between 2004 and 2007, and was set to increase further before the financial crisis of 2008. The main multilaterals--the African Development Bank, the European Commission, and the World Bank--secured record replenish- ment of concessional funding for three or four years from 2008. Attention to the possibility of reallocating ODA across countries would give impe- tus to the ongoing trilateral coordination of the African Development Bank, the World Bank, and the European Commission. While any discus- sion of a policy of redistribution of aid between countries would inevitably be politically charged, and might have to start at the level of the regional economic commissions, it could energize the prioritization of transport investments--the aim of the infrastructure plans of the New Partnership for Africa's Development and of the European Union-Africa Partnership on Infrastructure. Bilaterals may be more squeezed, and because of their smaller portfo- lios, less able to implement a policy of redistribution than multilaterals. Overall, however, it remains critical to the continued development of transport infrastructure in Africa that the scaling up of ODA is sustained. In sum, an agenda for African transport infrastructure finance might include the following steps: · Consider what is sustainable--avoid excessively high network densi- ties and design standards. Conclusion: An Agenda for Action 405 · Be cautious about "prestige projects"--that is where wasteful invest- ments thrive. · Invest only in what can be maintained--unsustainable investment is wasted. · Decide which parts of existing networks are sustainable, and maintain those properly. · Apply user fees to fund maintenance wherever possible. · Embrace low-cost ODA funds--but only when they can be invested to yield a high social rate of return. · Mobilize private capital for investment--but do not let it distort the investment program. · Supplement private capital with public sector funds in public-private partnership schemes in cases where this is the best way to maintain social service provision. That agenda for action is long and demanding. It requires changes in national institutions and attitudes as well as finance. So, while the inter- national community can help, success in bringing Africa's transport infrastructure up to the best international standards can be achieved (in keeping with the philosophy of the New Partnership for Africa's Development) only through African leadership and commitment. Notes 1. The Logistics Performance Index is based on a survey of worldwide operators (global freight forwarders and express carriers) providing feedback on seven characteristics of the logistics "friendliness" of 150 countries. 2. A port community system is a consultative association involving all principal stakeholders (public and private). Such systems have been very successful in some European ports, such as Rotterdam. 3. Development policy loans accounted for 50 percent of new lending by the World Bank in 2009 (World Bank 2009). References Bullock, R. 2009. "Railways in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 17, World Bank, Washington, DC. Drewry Shipping Consultants. 2009. Annual Review of Global Container Terminal Operators. London: Drewry Publications. South African Department of Transport. 2007. "National Freight Logistics Strategy." Department of Transport, Pretoria, South Africa. 406 Africa's Transport Infrastructure Transparency International. 2009. "Corruption Perceptions Index." Transparency International, Berlin. World Bank. 2007. Logistics Performance Index (software). World Bank, Washington, DC. ------. 2009. "Development Policy Lending Retrospective." World Bank. Washington, DC. ------. 2010. Logistics Performance Index (software). World Bank. Washington, DC. http://go.worldbank.org/88X6PU5GV0. APPENDIX 1 Introduction Appendix 1a AICD Background Documents Relevant to the Transport Sector The Africa Infrastructure Country Diagnostic (AICD) study covered all infrastructure sectors. In addition to the main summary report (Foster and Briceño-Garmendia 2009), the documentation included sets of background papers and working papers. The listing below (tables A1a.1 and A1a.2) includes both those that are specific to the transport sector and those that are of more general application. All papers are available for download from https://www.infrastructureafrica.org/aicd/. 407 408 Africa's Transport Infrastructure Table A1a.1 Background Papers No. Category and title Author(s) General 2 "Access, Affordability, and Alternatives: Sudeshna Banerjee, Quentin Wodon, Modern Infrastructure Services in Africa" Amadou Diallo, Taras Pushak, Helal Uddin, Clarence Tsimpo, and Vivien Foster 15 "Financing Public Infrastructure in Cecilia Briceño-Garmendia, Karlis Smits, Sub-Saharan Africa: Patterns, Issues, and Vivien Foster and Options" Investment needs 7 "Improving Connectivity: Investing Robin Carruthers and Ranga Rajan in Transport Infrastructure in Krishnamani with Siobhan Murray Sub-Saharan Africa" Subsector reviews 1 "Stuck in Traffic: Urban Transport in Africa" Ajay Kumar and Fanny Barrett 8 "Beyond the Bottlenecks: Ports in Mike Mundy and Andrew Penfold Sub-Saharan Africa" 11 "Railways in Sub-Saharan Africa" Richard Bullock 14 "The Burden of Maintenance: Roads in Ken Gwilliam, Vivien Foster, Rodrigo Sub-Saharan Africa" Archondo-Callao, Cecilia Briceño- Garmendia, Alberto Nogales, and Kavita Sethi 16 "Air Transport: Challenges to Growth" Heinrich C. Bofinger 17 "Taking Stock of Railway Companies in Mapapa Mbangala Sub-Saharan Africa" Source: Author's compilation. Table A1a.2 Working Papers No. Title Author(s) 1 "Making Sense of Sub-Saharan Africa's Tito Yepes, Justin Pierce, and Infrastructure Endowment: A Benchmarking Vivien Foster Approach" 3 "Infrastructure and Growth in Africa" César Calderón 8 "Potential for Local Private Finance Jacqueline Irving and Astrid Manroth of Infrastructure in Africa" 9 "Impact of Infrastructure Constraints on Alvaro Escribano, J. Luis Guasch, Firm Productivity in Africa" and Jorge Pena 10 "A Tale of Three Cities: Understanding Sumila Gulyani, Debabrata Differences in Provision of Modern Services" Talukdar, and Darby Jack 14 "Transport Prices and Costs in Africa: Supee Teravaninthorn and A Review of the Main International Corridors" Gaël Raballand (continued) Introduction 409 Table A1a.2 (continued) No. Title Author(s) 15 "The Impact of Infrastructure Spending in Jean-François Perrault and Sub-Saharan Africa: A CGE Modeling Luc Savard Approach" 17 "Fiscal Costs of Infrastructure Cecilia Briceño-Garmendia Provision: A Practitioner's Guide" 19 "Crop Production and Road Paul Dorosh, Hyoung-Gun Wang, Connectivity in Sub-Saharan Africa: Liang You, and Emily Schmidt A Spatial Analysis" Source: Author's compilation. Appendix 1b Country Typology for Study Countries Middle-income Resource-rich Low-income, Low-income, countries countries nonfragile countries fragile countries Other Botswana Angola Benin Burundi Djibouti Cape Verde Cameroon Burkina Faso Central African Republic Lesotho Chad Ethiopia Comoros Mauritius Congo, Rep. Ghana Congo, Dem. Rep. Namibia Equatorial Kenya Côte d'Ivoire Seychelles Guinea Madagascar Eritrea South Africa Gabon Malawi Gambia, The Swaziland Nigeria Mali Guinea Sudan Mauritania Guinea-Bissau Zambia Mozambique Liberia Niger São Tomé and Príncipe Rwanda Sierra Leone Senegal Togo Tanzania Zimbabwe Uganda Source: Foster and Briceño-Garmendia 2009. Reference Foster, V., and C. Briceño-Garmendia, eds. 2009. Africa's Infrastructure: A Time for Transformation. Paris: Agence Française de Développement; Washington, DC: World Bank. APPENDIX 2 Roads Appendix 2a Road Data Sources and Analysis Primary Data Sources The primary data on road infrastructure are drawn from three sources. The first is the Africa Infrastructure Country Diagnostic (AICD) road network survey. This includes an initial detailed survey of the nature, extent, and condition of road networks, performed in 21 of the 24 phase I AICD countries. In its second phase, the survey was extended to cover 40 countries. Where possible, statistics are based on the complete sample of 40; otherwise, the number of countries covered is specified. Fiscal data cover only the initial 21 countries surveyed. The AICD survey entailed consultant visits to the central road entity in each country to collect link- by-link information on the primary, secondary, and (when possible) terti- ary networks. For each network link, the survey ascertained the class (primary, secondary, or tertiary), the surface type (concrete, asphalt, gravel, or earth), the condition (good, fair, or poor), and the traffic vol- umes (across a series of five bands corresponding to typical values for each class of the network). The second data source is an institutional database maintained by the Sub-Saharan Africa Transport Policy Program (SSATP), which has tracked the development of institutional reforms in the African road sector in 411 412 Africa's Transport Infrastructure recent years, with emphasis on the design and adoption of road funds and road agencies. The third main source is the AICD fiscal costs study, which collected detailed data on road expenditures in the 24 phase I AICD countries. The data allow for disaggregating road fund and non­road fund expenditures; capital and operating expenditures; and in some cases, expenditures on the main network and those on the rural network. But it is not possible to capture the budget allocations that local jurisdictions make to their rural network, and as a result, rural network spending is almost certainly underrecorded, though to varying degrees. In addition to these three primary sources, chapter 2 draws extensively on work by the SSATP and on recent World Bank research on road freight transport operations in Sub-Saharan Africa. Throughout, the analysis dis- tinguishes between the main road network--that is, those parts of the net- work under the jurisdiction of the central road entity--and the rest of the system. In most countries, the main network includes both the primary and the secondary networks, but in a handful of cases (including larger countries such as Nigeria and South Africa) it comprises the primary net- work only. The rural network comprises the remainder of the classified network. This categorization is adopted so that data on historical road expenditures (which can be split only between the main and rural network and not among primary, secondary, and tertiary networks) can be reconciled with data on road network conditions and future expenditure needs. Due to the lower quality of the available data on both rural network conditions and rural network expenditure, the analysis of the rural networks is neces- sarily more speculative than that of the main networks. Methodology of the RONET Model The Road Network Evaluation Tools (RONET) model is a tool for assess- ing the performance of road maintenance and rehabilitation policies and the importance of the road sector to the economy. The model is used to demonstrate to stakeholders the need for continued support of road maintenance initiatives. The length of the road network taken into con- sideration in the model may be the entire road system of the country (roads, highways, expressways, streets, avenues, and so forth), or a partial network--for example, the road network of a state or province of the country or the road network managed by the main road agency. Segments of the road network are classified according to (i) five network types, (ii) five surface types, (iii) five traffic categories, and (iv) five condition categories, for a total of 625 road classes (figure A2a.1). Roads 413 Figure A2a.1 Matrix of Road Classes: Overall Network Evaluation network surface type type concrete asphalt ST gravel earth primary secondary tertiary unclassified urban traffic condition category category very good good fair poor very poor traffic I traffic II traffic III traffic IV traffic V Source: Archondo Callao 2009. Note: ST = surface treatment. Each surface type is subdivided into five possible traffic categories. Table A2a.1 presents the default traffic level assigned to each combina- tion of traffic category and surface type. Each network type, road type, and traffic category is further subdivided into five possible road condi- tion categories defined as a function of the engineering assessment of the capital road works (periodic maintenance or rehabilitation works) needed to bring a road to very good condition. Routine maintenance road works are needed by all roads every year; therefore, they are not considered in the definitions of the road condition classes. The road con- dition classes are defined as follows: Very good: Roads in very good condition require no capital road works. Good: Roads in good condition are largely free of defects and require only some minor maintenance works, such as preventive treatment, crack sealing, or grading. Fair: Roads in fair condition are roads with defects and weakened structural resistance, requiring resurfacing of the pavement (periodic maintenance) but not demolition of the existing pavement. Poor: Roads in poor condition require rehabilitation (strengthening or partial reconstruction). 414 Table A2a.1 RONET Default Assignment of Traffic Levels Average annual daily traffic (vehicles per day) Illustrative standards Surface type Traffic category Traffic level Minimum Maximum Average Geometry Pavement Earth Traffic I T1 0 10 5 1-lane warranted Formation not warranted Traffic II T2 10 30 20 1-lane warranted Formation warranted Traffic III T3 30 100 65 2-lane warranted Gravel warranted Traffic IV T4 100 300 200 2-lane warranted Gravel warranted Traffic V T5 300 1,000 650 2-lane warranted Paved surface warranted Gravel Traffic I T2 10 30 20 1-lane warranted Formation warranted Traffic II T3 30 100 65 2-lane warranted Gravel warranted Traffic III T4 100 300 200 2-lane warranted Gravel warranted Traffic IV T5 300 1,000 650 2-lane warranted Paved surface warranted Traffic V T6 1,000 3,000 2,000 2-lane warranted Paved surface warranted Paved Traffic I T4 100 300 200 2-lane warranted Gravel warranted Traffic II T5 300 1,000 650 2-lane warranted Paved surface warranted Traffic III T6 1,000 3,000 2,000 2-lane warranted Paved surface warranted Traffic IV T7 3,000 10,000 6,500 2-lane warranted Paved surface warranted Traffic V T8 10,000 30,000 20,000 4-lane warranted Paved surface warranted Source: Archondo Callao 2009. Note: Standard given for illustrative purposes. Proper standards are country specific. Average annual daily traffic is for motorized vehicles with four wheels or more in two-way traffic. Roads 415 Very poor: Roads in very poor condition require full reconstruction, almost equivalent to new construction. RONET has one module that computes the network monitoring indi- cators based on the current condition of the network and another that does a performance assessment of the network under different road agency standards. The objective of the latter is to assess the consequences of applying different standards that represent different levels of expendi- tures on road works over time. The consequences are reflected in the road works requirements, including financial cost, road condition, asset value, and so on. This module evaluates the performance of the network under differ- ent road works standards over a 20-year evaluation period. The standards that may be selected by users of the model are the following: · Very high standard, which represents a scenario without budget con- straints but with a high level of periodic maintenance and rehabilita- tion works · High, medium, low, and very low standards, which represent scenarios of decreasing levels of road works expenditures · Do-the-minimum standard, in which the only capital road work applied over the evaluation period is reconstruction at a very high roughness. · Do-nothing standard, in which no capital road works are applied over the evaluation period ·Custom standard, in which one of the above standards is applied indi- vidually to each road network type · Optimal standard, for which RONET evaluates each road class and identifies the standard that maximizes the net present value of social benefits at a given discount rate. RONET was used to evaluate the preservation needs of the primary and secondary roads of 19 African countries. The total network length of the 19 countries is 991,567 kilometers (km), of which 293,039 km (30 percent) are main roads. The total network utilization of the 19 countries is 123,755 million vehicle-km, of which 117,905 million vehicle-km (95 percent) circulate on main roads. The median traffic vol- ume on the main roads is 456 vehicles per day. RONET evaluated the cur- rent condition and traffic of the main roads and computed current monitoring indicators for each country. The median percentage of roads in good and fair condition is 75.6 percent, and the median average network roughness weighted per vehicle-km is 4.69 meters per kilometer. This is 416 Africa's Transport Infrastructure referred to as the International Roughness Index (IRI). The median cur- rent asset value as a share of the maximum asset value is 85 percent, and the median current asset value as a share of gross domestic product (GDP) is 17 percent. RONET evaluated the performance of the main roads under different preservation standards to determine the optimal needs for recurrent maintenance, periodic maintenance, and rehabilitation road works. For each road class--characterized by functional classification, surface type, traffic, and condition--RONET identified the preservation standard that yields the highest net present value at a 12 percent discount rate and thus maximizes society's net benefits. The RONET evaluation assumed the following in all countries: (i) a traffic growth rate of 3 percent per year, (ii) discount rate of 12 percent, (iii) evaluation period of 20 years, (iv) average unit costs of road works based on the World Bank Road Works Costs Knowledge System, and (v) average unit road user costs based on the World Bank Road User Costs Knowledge System. Appendix 2b Basic Country Data for the Set of 40 Countries GDP Vehicle current fleet Land Population (millions) prices (number of Country area (km2) Total Rural Urban (US$ billions) vehicles) Angola 1,246,700 18.0 7.8 10.2 83.4 691,192 Benin 110,620 8.7 5.1 3.6 6.7 229,536 Botswana 566,730 1.9 0.8 1.1 13.0 302,568 Burkina Faso 273,600 15.2 12.2 3.0 7.9 530,917 Burundi 25,680 8.1 7.2 0.8 1.2 61,271 Cameroon 465,400 18.9 8.2 10.7 23.4 31,627 Central African Republic 623,000 4.4 2.7 1.7 2.0 18,540 Chad 1,259,200 11.1 8.1 3.0 8.4 127,811 Congo, Dem. Rep. 2,267,050 64.2 42.4 21.8 11.6 321,134 Congo, Rep. 341,500 3.6 1.4 2.2 10.7 103,000 Côte d'lvoire 318,000 20.6 10.5 10.0 23.4 362,560 Eritrea 101,000 5.0 4.0 1.0 1.7 62,674 Ethiopia 1,000,000 80.7 67.0 13.7 26.5 251,585 Gabon 257,670 1.4 0.2 1.2 14.4 20,600 Gambia, The 10,000 1.7 0.7 0.9 0.8 14,884 Ghana 227,540 23.4 11.7 11.7 16.1 959,591 Guinea 245,720 9.8 6.4 3.4 4.3 154,500 Kenya 569,140 38.5 30.2 8.3 34.5 1,034,370 (continued) Roads 417 GDP Vehicle current fleet Land Population (millions) prices (number of Country area (km2) Total Rural Urban (US$ billions) vehicles) Lesotho 30,350 2.0 1.5 0.5 1.6 77,217 Liberia 96,320 3.8 1.5 2.3 0.9 11,419 Madagascar 581,540 19.1 13.5 5.6 9.0 203,920 Malawi 94,080 14.3 11.6 2.7 4.3 133,900 Mali 1,220,190 12.7 8.6 4.1 8.7 172,262 Mauritania 1,030,700 3.2 1.9 1.3 2.9 360,500 Mauritius 2,030 1.3 0.7 0.5 8.7 344,149 Mozambique 786,380 21.8 13.8 8.0 9.7 266,440 Namibia 823,290 2.1 1.3 0.8 8.6 246,800 Niger 1,266,700 14.7 12.2 2.4 5.4 78,343 Nigeria 910,770 151.3 78.1 73.2 212.1 7,828,000 Rwanda 24,670 9.7 7.9 1.8 4.5 62,830 Senegal 192,530 12.2 7.0 5.2 13.2 289,012 Sierra Leone 71,620 5.6 3.5 2.1 2.0 46,359 South Africa 1,214,470 48.7 19.1 29.6 276.8 9,514,701 Sudan 2,376,000 41.3 23.4 18.0 58.4 1,236,000 Swaziland 17,200 1.2 0.9 0.3 2.6 119,532 Tanzania 885,800 42.5 31.6 10.8 20.5 595,287 Togo 54,390 6.5 3.7 2.7 2.8 49,681 Uganda 197,100 31.7 27.5 4.1 14.5 374,568 Zambia 743,390 12.6 8.2 4.5 14.3 228,854 Zimbabwe 386,850 12.5 7.8 4.7 3.4 1,603,284 Total 22,914,920 805.8 512.2 293.6 974.559 29,121,417 Average 572,873 20.1 12.8 7.3 24.364 728,035 Median 364,175 12.3 7.9 3.2 8.696 229,195 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. Note: GDP = gross domestic product; km2 = square kilometer. 418 Appendix 2c Classified Road Network Length for 40 Countries (kilometers) Classified road network Total including Grand total Country Primary Secondary Tertiary Classified Unclassified unclassified Urban including urban Angola 10,227 11,069 15,104 36,399 15,104 51,503 11,057 62,560 Benin 2,386 2,348 3,597 8,332 7,368 15,700 1,794 17,494 Botswana 4,028 4,705 9,002 17,735 5,668 23,403 4,749 28,152 Burkina Faso 6,705 3,526 4,971 15,202 7,108 22,310 1,889 24,199 Burundi 1,839 1,126 1,113 4,077 5,538 9,615 462 10,077 Cameroon 5,838 5,169 12,573 23,581 9,802 33,383 5,772 39,155 Central African Republic 5,037 4,952 7,623 17,612 7,623 25,235 3,544 28,779 Chad 7,064 5,891 15,283 28,238 5,162 33,400 3,202 36,602 Congo, Dem. Rep 19,451 8,850 5,634 33,934 5,238 39,172 9,123 48,295 Congo, Rep. 1,619 3,786 7,330 12,735 4,174 16,909 2,939 19,848 Côte d'lvoire 5,752 7,539 12,267 25,558 516 26,074 5,887 31,961 Eritrea 2,171 2,241 1,400 5,812 2,373 8,185 1,190 9,375 Ethiopia 4,953 4,768 10,906 20,627 25,081 45,708 1,521 47,229 Gabon 2,954 1,713 4,317 8,984 3,100 12,083 3,733 15,816 Gambia, The 812 847 1,300 2,959 790 3,749 858 4,607 Ghana 3,564 7,613 17,452 28,628 11,634 40,263 5,574 45,837 Guinea 3,719 2,642 9,405 15,767 7,240 23,007 2,043 25,050 Kenya 5,532 16,809 34,609 56,951 6,314 63,265 4,237 67,502 Lesotho 1,483 1,521 2,295 5,299 641 5,940 702 6,642 Liberia 2,378 2,160 4,629 9,167 7,427 16,594 2,518 19,112 Madagascar 2,991 6,608 16,042 25,641 3,952 29,593 2,517 32,110 Malawi 3,444 6,692 3,147 13,283 2,208 15,491 1,702 17,193 Mali 7,017 11,203 15,852 34,072 12,667 46,739 2,258 48,997 Mauritania 5,303 1,796 2,831 9,930 2,505 12,435 2,791 15,226 Mauritius 896 592 527 2,015 42 2,057 591 2,648 Mozambique 4,909 4,900 12,689 22,498 6,740 29,238 3,321 32,559 Namibia 4,297 12,065 28,732 45,094 18,067 63,161 5,307 68,468 Niger 4,011 2,044 7,371 13,427 3,518 16,945 947 17,892 Nigeria 28,614 21,293 35,900 85,807 72,800 158,607 22,465 181,072 Rwanda 1,060 1,776 1,790 4,625 9,383 14,008 1,988 15,996 Senegal 2,920 1,860 10,891 15,671 2,392 18,063 1,898 19,961 Sierra Leone 2,312 2,091 4,152 8,555 444 8,999 3,000 11,999 South Africa 38,066 38,222 125,975 202,263 161,868 364,131 42,572 406,703 Sudan 10,122 6,682 8,199 25,003 5,542 30,545 6,829 37,374 Swaziland 1,467 1,647 1,192 4,306 569 4,875 388 5,263 Tanzania 7,010 21,720 20,000 48,730 5,727 54,457 6,433 60,890 Togo 1,852 992 1,834 4,678 2,842 7,520 1,127 8,647 Uganda 9,171 26,751 35,000 70,922 5,000 75,922 4,100 80,022 Zambia 3,290 3,763 11,232 18,285 18,958 37,243 2,282 39,525 Zimbabwe 5,270 6,758 7,772 19,800 18,984 38,784 5,000 43,784 Total 241,536 278,730 531,938 1,052,203 492,109 1,544,312 190,311 1,734,623 Average 6,038 6,968 13,298 26,305 12,303 38,608 4,758 43,366 Median 4,020 4,736 7,986 17,674 5,605 24,319 2,865 28,465 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. 419 Appendix 2d Road Network Densities for 40 Countries 420 Classified network length density Total network length density Network density per population Per area Per population Per vehicles Per area Per population Per vehicles Tertiary network Urban network (km/1,000 (km/1,000 (km/1,000 (km/1,000 (km/1,000 (km/1,000 (km/1,000 (km/1,000 Country km2) people) vehicles) km2) people) vehicles) rural people) urban people) Angola 29 2.0 53 41 2.9 53 1.9 1.1 Benin 75 1.0 36 142 1.8 36 0.7 0.5 Botswana 31 9.3 59 41 12.3 59 11.7 4.2 Burkina Faso 56 1.0 29 82 1.5 29 0.4 0.6 Burundi 159 0.5 67 374 1.2 67 0.2 0.6 Cameroon 51 1.2 746 72 1.8 746 1.5 0.5 Central African Republic 28 4.0 950 41 5.7 950 2.8 2.1 Chad 22 2.6 221 27 3.0 221 1.9 1.1 Congo, Dem. Rep. 15 0.5 106 17 0.6 106 0.1 0.4 Congo, Rep. 37 3.5 124 50 4.7 124 5.2 1.3 Côte d'lvoire 80 1.2 70 82 1.3 70 1.2 0.6 Eritrea 58 1.2 93 81 1.6 93 0.4 1.1 Ethiopia 21 0.3 82 46 0.6 82 0.2 0.1 Gabon 35 6.2 436 47 8.3 436 19.9 3.0 Gambia, The 296 1.8 199 375 2.3 199 1.8 0.9 Ghana 126 1.2 30 177 1.7 30 1.5 0.5 Guinea 64 1.6 102 94 2.3 102 1.5 0.6 Kenya 100 1.5 55 111 1.6 55 1.1 0.5 Lesotho 175 2.6 69 196 2.9 69 1.5 1.4 Liberia 95 2.4 803 172 4.4 803 3.1 1.1 Madagascar 44 1.3 126 51 1.5 126 1.2 0.4 Malawi 141 0.9 99 165 1.1 99 0.3 0.6 Mali 28 2.7 198 38 3.7 198 1.8 0.6 Mauritania 10 3.1 28 12 3.9 28 1.5 2.1 Mauritius 993 1.6 6 1,014 1.6 6 0.7 1.1 Mozambique 29 1.0 84 37 1.3 84 0.9 0.4 Namibia 55 21.3 183 77 29.9 183 21.5 6.8 Niger 11 0.9 171 13 1.2 171 0.6 0.4 Nigeria 94 0.6 11 174 1.0 11 0.5 0.3 Rwanda 187 0.5 74 568 1.4 74 0.2 1.1 Senegal 81 1.3 54 94 1.5 54 1.5 0.4 Sierra Leone 119 1.5 185 126 1.6 185 1.2 1.4 South Africa 167 4.2 21 300 7.5 21 6.6 1.4 Sudan 11 0.6 20 13 0.7 20 0.4 0.4 Swaziland 250 3.7 36 283 4.2 36 1.4 1.3 Tanzania 55 1.1 82 61 1.3 82 0.6 0.6 Togo 86 0.7 94 138 1.2 94 0.5 0.4 Uganda 360 2.2 189 385 2.4 189 1.3 1.0 Zambia 25 1.4 80 50 3.0 80 1.4 0.5 Zimbabwe 51 1.6 12 100 3.1 12 1.0 1.1 Average 109 2.5 152 149 3.4 152 2.6 1.1 Median 57 1.5 82 82 1.7 82 1.2 0.6 Source: Calculations by Alberto Nogales based on AICD RONET Summary Outputs, June 2010. Note: km = kilometer; km2 = square kilometer. 421 422 Appendix 2e Road Network Length by Surface Class and Network Type for 40 Countries (kilometers) Paved road network Unpaved road network Country Primary Secondary Tertiary Classified Primary Secondary Tertiary Classified Angola 7,383 1,235 0 8,618 2,843 9,835 15,104 27,781 Benin 1,821 0 0 1,821 565 2,348 3,597 6,510 Botswana 4,028 2,268 645 6,942 0 2,436 8,356 10,793 Burkina Faso 2,640 23 7 2,670 4,065 3,502 4,965 12,532 Burundi 1,059 0 2 1,061 780 1,126 1,111 3,016 Cameroon 3,113 903 523 4,540 2,725 4,266 12,050 19,041 Central African Republic 654 0 0 654 4,383 4,952 7,623 16,958 Chad 908 78 0 986 6,156 5,813 15,283 27,252 Congo, Dem. Rep 2,269 0 4 2,274 17,181 8,850 5,629 31,661 Congo, Rep. 541 468 42 1,051 1,079 3,318 7,288 11,684 Côte d'lvoire 3,973 1,643 73 5,689 1,779 5,896 12,194 19,869 Eritrea 817 28 0 845 1,354 2,214 1,400 4,968 Ethiopia 4,132 255 311 4,698 821 4,513 10,595 15,928 Gabon 941 164 34 1,139 2,013 1,549 4,283 7,845 Gambia, The 812 0 0 812 0 847 1,300 2,147 Ghana 2,204 3,157 753 6,114 1,360 4,456 16,699 22,514 Guinea 2,092 235 0 2,327 1,627 2,407 9,405 13,440 Kenya 3,787 3,470 725 7,982 1,744 13,339 33,885 48,968 Lesotho 847 191 3 1,041 636 1,330 2,292 4,258 Liberia 656 2 101 758 1,722 2,158 4,528 8,408 Madagascar 2,537 2,142 366 5,045 454 4,466 15,676 20,596 Malawi 2,486 423 96 3,004 958 6,270 3,051 10,278 Mali 3,873 602 71 4,545 3,145 10,601 15,782 29,527 Mauritania 1,900 37 14 1,951 3,403 1,759 2,816 7,979 Mauritius 896 592 527 2,015 0 0 0 0 Mozambique 4,360 880 458 5,698 549 4,020 12,231 16,800 Namibia 4,177 1,622 232 6,031 120 10,443 28,499 39,062 Niger 3,549 69 32 3,649 462 1,975 7,340 9,777 Nigeria 24,505 13,506 787 38,798 4,109 7,787 35,113 47,009 Rwanda 1,060 0 0 1,060 0 1,776 1,790 3,566 Senegal 2,700 1,191 189 4,080 219 669 10,702 11,591 Sierra Leone 905 46 0 951 1,407 2,045 4,152 7,604 South Africa 37,876 23,720 12,971 74,567 190 14,502 113,004 127,697 Sudan 2,068 161 0 2,229 8,055 6,520 8,199 22,775 Swaziland 1,043 80 6 1,129 424 1,567 1,186 3,177 Tanzania 3,297 1,033 0 4,330 3,713 20,687 20,000 44,400 Togo 1,591 15 0 1,606 261 977 1,834 3,071 Uganda 2,331 0 0 2,331 6,840 26,751 35,000 68,591 Zambia 3,290 2,058 1,111 6,459 0 1,705 10,121 11,826 Zimbabwe 5,270 2,652 0 7,922 0 4,106 7,772 11,878 Total 154,392 64,947 20,084 239,424 87,143 213,782 511,854 812,779 Average 3,860 1,624 502 5,986 2,179 5,345 12,796 20,319 Median 2,300 245 23 2,500 1,216 3,761 7,986 12,205 Source: Calculations by Alberto Nogales based on AICD RONET Summary Outputs, June 2010. 423 424 Africa's Transport Infrastructure Appendix 2f Average Annual Daily Traffic by Road Type for 40 Countries Road type Country Primary Secondary Tertiary Classified Angola 733 28 6 136 Benin 1,627 82 15 243 Botswana 1,381 386 41 284 Burkina Faso 401 49 30 130 Burundi 306 11 6 64 Cameroon 1,012 263 34 204 Central African Republic 55 26 6 23 Chad 159 35 6 44 Congo, Dem. Rep. 48 19 19 34 Congo, Rep. 251 93 24 55 Côte d'lvoire 788 235 18 212 Eritrea 397 46 6 111 Ethiopia 555 153 92 100 Gabon 315 43 8 77 Gambia, The 375 31 6 83 Ghana 1,917 285 35 385 Guinea 635 140 6 117 Kenya 1,306 233 26 182 Lesotho 1,789 567 70 559 Liberia 317 33 10 54 Madagascar 937 299 13 160 Malawi 628 77 32 167 Mali 326 75 8 70 Mauritania 253 10 7 100 Mauritius 7,482 600 200 2,750 Mozambique 1,053 163 73 224 Namibia 1,644 221 31 160 Niger 415 64 33 118 Nigeria 2,310 746 40 469 Rwanda 913 125 6 84 Senegal 1,286 431 42 256 Sierra Leone 503 75 7 122 South Africa 6,596 126 26 676 Sudan 284 24 13 93 Swaziland 1,997 299 7 656 Tanzania 1,276 247 13 244 Togo 2,054 666 8 526 Uganda 870 37 13 121 Zambia 1,424 203 56 163 Zimbabwe 1,294 161 13 190 Average 1,198 185 28 261 Median 829 126 14 148 Source: Calculations by Alberto Nogales based on AICD RONET Summary Outputs, June 2010. Appendix 2g Distribution of Networks by Traffic Level for 40 Countries (percent) T1 T2 T3 T4 T5 T6 T7 T8 0­10 10­30 30­100 100­300 300­1,000 1,000­3,000 3,000­10,000 >10,000 Country AADT AADT AADT AADT AADT AADT AADT AADT Angola 4 0 0 18 18 42 25 0 Benin 2 0 1 8 13 35 33 12 Botswana 1 0 1 21 22 23 35 0 Burkina Faso 4 8 12 9 26 47 0 0 Burundi 12 0 0 22 54 12 12 0 Cameroon 2 0 1 14 28 30 29 0 Central African Republic 55 22 8 60 0 0 0 0 Chad 14 3 29 27 33 6 0 0 Congo, Dem. Rep. 58 0 0 63 16 0 0 0 Congo, Rep. 4 29 0 13 66 5 0 0 Côte d'Ivoire 2 0 4 14 28 50 8 0 Eritrea 4 2 23 11 18 48 0 0 Ethiopia 8 1 12 25 40 18 0 0 Gabon 8 0 0 19 37 54 0 0 Gambia, The 10 0 3 41 47 12 0 0 Ghana 3 5 9 4 16 52 94 0 Guinea 4 2 8 12 30 25 6 18 Kenya 4 3 5 11 12 36 31 0 Lesotho 0 2 5 8 14 5 24 43 Liberia 10 13 12 19 16 48 0 0 425 Madagascar 3 4 3 12 23 11 21 25 (continued) 426 T1 T2 T3 T4 T5 T6 T7 T8 0­10 10­30 30­100 100­300 300­1,000 1,000­3,000 3,000­10,000 >10,000 Country AADT AADT AADT AADT AADT AADT AADT AADT Malawi 1 11 7 12 31 41 0 0 Mali 9 3 6 20 42 22 2 0 Mauritania 6 0 0 31 23 50 0 0 Mauritius 0 0 0 3 5 5 50 39 Mozambique 1 9 5 11 12 55 10 2 Namibia 2 8 6 8 13 14 41 11 Niger 4 5 7 23 49 14 0 0 Nigeria 1 0 2 3 12 33 35 15 Rwanda 7 0 1 23 25 55 0 0 Senegal 2 0 1 11 24 45 19 0 Sierra Leone 6 5 3 23 4 10 57 0 South Africa 1 5 0 3 2 7 29 59 Sudan 5 15 1 16 14 60 0 0 Swaziland 0 0 6 12 6 27 43 5 Tanzania 1 6 11 13 15 17 26 14 Togo 1 0 1 7 23 24 0 46 Uganda 3 7 10 13 15 13 32 10 Zambia 1 7 2 18 14 16 41 6 Zimbabwe 3 1 2 12 21 41 24 0 Average 7 4 5 17 23 28 18 8 Median 3 2 3 13 20 24 11 0 Source: Calculations by Alberto Nogales based on AICD RONET Summary Outputs, June 2010. Note: The columns cover ranges of AADT (average annual daily traffic), increasing from T1 to T8, which comprehensively cover the whole range of flows. Each row thus totals 100 percent (subject to rounding error). Roads 427 Appendix 2h Vehicle Utilization of Roads by Surface Class and by Passenger and Freight for 40 Countries Vehicle utilization in vehicle-km (millions) Country Primary Secondary Tertiary Classified Angola 2,737 113 33 2,882 Benin 1,417 71 20 1,507 Botswana 2,030 662 133 2,826 Burkina Faso 981 63 54 1,098 Burundi 205 4 3 212 Cameroon 2,156 496 154 2,806 Central African Republic 102 47 17 165 Chad 410 74 34 518 Congo, Dem. Rep. 338 61 40 439 Congo, Rep. 148 128 64 340 Côte d'lvoire 1,654 647 80 2,381 Eritrea 315 38 3 356 Ethiopia 1,002 267 367 1,637 Gabon 340 27 12 379 Gambia, The 111 10 3 124 Ghana 2,494 793 224 3,511 Guinea 862 135 21 1,018 Kenya 2,637 1,429 329 4,396 Lesotho 969 315 58 1,342 Liberia 275 26 17 318 Madagascar 1,023 720 78 1,821 Malawi 789 187 37 1,013 Mali 835 307 47 1,189 Mauritania 490 7 7 504 Mauritius 2,447 130 38 2,615 Mozambique 1,887 292 340 2,518 Namibia 2,579 972 329 3,880 Niger 607 48 90 745 Nigeria 24,123 5,794 522 30,439 Rwanda 353 81 4 438 Senegal 1,371 292 169 1,832 Sierra Leone 424 57 11 492 South Africa 91,648 1,763 1,194 94,605 Sudan 1,049 60 40 1,149 Swaziland 1,069 180 3 1,252 Tanzania 3,265 1,954 95 5,314 Togo 1,389 241 5 1,635 Uganda 2,914 363 166 3,443 Zambia 1,710 279 231 2,220 Zimbabwe 2,489 396 36 2,921 Total 163,644 19,530 5,106 188,280 Average 4,091 488 128 4,707 Median 1,036 183 43 1,425 Source: Calculations by Alberto Nogales based on AICD RONET Summary Outputs, June 2010. 428 Appendix 2i Classified Road Network Condition by Network Type for 40 Countries (percent) Primary network Secondary network Tertiary network Classified network Country Good Fair Poor Good Fair Poor Good Fair Poor Good Fair Poor Angola 42 30 28 23 21 55 20 20 60 27 23 50 Benin 33 25 43 7 74 18 23 43 34 21 46 32 Botswana 68 22 10 68 16 16 31 34 35 49 27 24 Burkina Faso 75 21 4 72 11 17 63 28 9 70 21 9 Burundi 50 24 26 40 20 40 40 20 40 44 22 34 Cameroon 42 26 32 39 28 34 32 26 41 36 27 37 Central African Republic 60 19 20 56 16 29 40 20 40 50 19 31 Chad 35 43 22 28 27 44 39 20 40 36 27 37 Congo, Dem. Rep. 14 15 71 31 17 51 30 21 49 21 17 62 Congo, Rep. 4 23 73 4 2 94 24 12 65 15 10 74 Côte d'Ivoire 16 63 21 4 91 5 29 30 41 18 56 26 Eritrea 32 30 37 5 35 61 40 20 40 24 29 47 Ethiopia 61 25 15 28 38 34 32 25 43 38 28 34 Gabon 40 20 40 40 20 40 40 20 40 40 20 40 Gambia, The 20 40 40 20 20 60 20 25 55 20 28 52 Ghana 45 37 18 41 29 30 51 36 13 48 34 18 Guinea 19 21 60 8 10 82 40 20 40 30 18 52 Kenya 51 31 18 51 23 26 49 11 40 50 16 34 Lesotho 19 39 42 25 32 43 31 24 45 26 30 43 Liberia 21 61 19 1 35 64 40 20 40 26 34 40 Madagascar 64 22 14 25 31 44 10 5 85 20 14 66 Malawi 43 47 11 39 51 11 44 42 14 41 48 12 Mali 45 29 26 34 31 35 39 22 39 39 26 35 Mauritania 41 38 21 42 19 39 40 19 41 41 29 30 Mauritius 77 20 3 77 20 3 77 20 3 77 20 3 Mozambique 58 30 12 26 52 21 21 35 44 30 38 32 Namibia 56 36 8 44 42 14 33 25 43 38 30 32 Niger 47 42 12 33 33 34 31 24 45 36 31 33 Nigeria 38 37 24 14 38 48 31 33 36 29 36 35 Rwanda 41 48 11 7 21 73 0 0 100 12 19 69 Senegal 33 12 55 26 26 48 21 18 61 24 18 58 Sierra Leone 41 16 42 32 31 36 34 18 48 36 21 44 South Africa 90 8 2 31 34 35 31 24 45 42 23 35 Sudan 31 34 35 21 20 59 20 22 58 25 27 49 Swaziland 71 22 7 47 41 13 40 20 40 53 29 18 Tanzania 59 32 9 30 51 20 31 24 45 34 37 29 Togo 23 6 70 0 1 99 40 20 40 25 11 65 Uganda 14 73 13 0 11 89 40 20 40 22 23 55 Zambia 55 16 29 31 21 48 16 30 54 26 26 48 Zimbabwe 30 50 20 24 26 50 25 29 47 26 33 41 Average 43 31 27 29 29 42 33 23 43 34 27 39 Median 41 29 21 29 26 39 32 21 41 32 27 36 Source: Calculations by Alberto Nogales based on AICD RONET Summary Outputs, June 2010. Note: For each network category, the percentages for the three road conditions add to 100 percent (subject to rounding error). 429 430 Africa's Transport Infrastructure Appendix 2j Road Accident Rates for Countries in Africa In 2009, the World Health Organization published a study entitled "Global Status Report on Road Safety: Time for Action." It was based on a self-administered survey on road safety undertaken in 172 participating countries in 2008. For each country, a national data coordinator, assisted by a team of six to eight key respondents, prepared and checked the sur- vey response. This included statistical data on the number of reported deaths and injury accidents, as well as institutional and administrative data on the approach to road safety in the country. Forty African coun- tries were included in the survey. The data collected included recorded deaths, population, number of registered vehicles, and average per capita income (see table A2j.1). The survey results showed the great difficulties of determining the road safety situation in Africa. Sixteen of the countries, while presenting statistics for deaths in 2008, were unable to give statistics for the trend in accidents. While injury accidents are always notoriously difficult to record and assess in all countries, the vast majority of African countries were also judged to underrecord deaths, in many cases by a factor of 3 or 4 and in one case by a factor of 10. The World Health Organization adjusted the raw fatalities data in two ways. First, using the factors adopted by the European Conference of Ministers of Transport to secure consistency for a 30-day period, an adjustment was made to secure consistency in the definition of a traffic- related fatality. Second, information on the completeness of reporting was obtained from previous World Health Organization surveys. For countries with a high reporting level, a negative binomial model was calibrated explaining the death rates as a function of a set of independent variables described as exposure factors (Ej), risk or preventive factors (Rj), mitigat- ing factors (Mj), and national income (Ij). Thus, Yj = f(Rj, Mj, Ej, Ij). Despite the efforts to adjust for underrecording, the figures are still problematic. The estimated death rates per 10,000 people published in the report showed a very narrow range, between 23 in Burundi and 48 in Eritrea. Most values fell in the high 20s or low 30s. That might not be too strange, as the countries share the characteristics of relatively low income and poor policy provision for accident prevention; however, applying the same number of deaths to the registered vehicle populations in the vari- ous countries showed fatality rates per vehicle ranging from just over 2 to over 200 per 1,000 vehicles. Such a pattern of variation is hardly credible Table A2j.1 Reported and Estimated Deaths from Road Accidents, 2008 Death rate Population GNP per Number of Reported Estimated per 1,000 Death rate per Country 2007 capita registered vehicles deaths deaths population 1,000 vehicles Angola 17,024,084 2,560 671,060 2,358 6,425 38 9.5744047 Benin 9,032,787 570 222,850 653 2,815 31 12.631815 Botswana 1,881,504 5,840 293,755 482 636 34 2.1650695 Burkina Faso 14,784,291 430 515,453 804 4,595 31 8.9144888 Burundi 8,508,232 110 59,486 63 1,989 23 33.436439 Cameroon 18,549,176 1,050 312,259 1,069 5,206 28 16.672057 Cape Verde 530,437 2,430 54,158 49 133 25 2.4557775 Central African Republic 4,342,735 380 5,834 569 1,399 32 239.80117 Chad 10,780,571 540 124,088 814 3,696 34 29.785314 Congo, Rep. 3,768,086 1,540 100,000 207 1,084 29 10.84 Congo, Dem. Rep. 62,635,723 140 311,781 365 20,183 32 64.734541 Eritrea 4,850,763 230 60,849 81 2,350 48 38.620191 Ethiopia 83,099,190 220 244,257 2,441 29,114 35 119.19413 Gambia 1,708,601 320 14,450 54 625 37 43.252595 Ghana 23,478,394 590 931,642 1,856 6,942 30 7.4513601 Guinea-Bissau 1,695,043 200 57,839 152 583 34 10.079704 Kenya 37,537,716 680 1,004,243 3,760 12,198 34 12.146463 Lesotho 2,007,833 1,000 -- 402 537 27 -- Liberia 3,750,261 150 11,086 -- 1,235 33 111.40177 Madagascar 19,683,358 320 197,981 594 6,641 34 33.543623 Malawi 13,925,070 250 130,000 839 3,614 26 27.8 431 (continued) 432 Table A2j.1 (continued) Death rate Population GNP per Number of Reported Estimated per 1,000 Death rate per Country 2007 capita registered vehicles deaths deaths population 1,000 vehicles Mali 12,336,799 500 167,245 711 3,959 32 23.671859 Mauritania 3,123,813 840 350,000 262 1,109 36 3.1685714 Mozambique 21,396,916 320 258,680 1,952 7,432 35 28.730478 Namibia 2,074,146 3,360 239,612 368 594 29 2.4790077 Niger 14,225,521 280 76,061 570 5,357 38 70.430313 Nigeria 148,092,542 930 7,600,000 4,532 47,865 32 6.2980263 Rwanda 9,724,577 320 61,000 308 3,077 32 50.442623 São Tomé and Príncipe 157,638 870 1,219 20 52 33 42.657916 Senegal 12,378,532 820 280,594 345 4,023 33 14.337441 Sierra Leone 5,865,872 260 39,038 68 1,661 28 42.548286 South Africa 48,576,763 5,760 9,237,574 16,113 16,113 33 1.7442891 Sudan 38,560,488 960 1,200,000 2,277 13,362 35 11.135 Swaziland 1,141,427 2,580 116,050 235 300 26 2.5850926 Togo 6,585,147 360 48,234 613 1,851 28 38.37542 Uganda 30,883,805 340 3,673,658 2,838 7,634 25 2.0780377 Zambia 11,921,999 800 222,188 1,645 3,056 26 13.754118 Zimbabwe 13,349,434 131 1,556,586 1,348 3,669 28 2.3570815 Source: World Health Organization 2009. Note: -- = Not available. Roads 433 and probably reflects a number of factors, including underrecording of vehicle numbers in extreme cases (such as the Central African Republic). Despite misgivings about the spread of results, some fairly strong con- clusions can be drawn. First, the average death rates in Africa are very high in comparison with world averages, matched only by rates in parts of South Asia. Second, where trends were recorded over time, death rates were rising in 13 countries, variable or static in 9, and falling in only 2. This pattern is contrary to the experience in most parts of the world, even in countries where vehicle stock is increasing. 434 Appendix 2k Road Maintenance Initiative Institutional Indicators, September 2007 Share of road Share Coverage Transport Long-term road Board fund of road user of routine Coverage policya investment program with Staff resources charges maintenance of total Fuel levye Cabinet Adoption Period Date of Has a private size from user from fuel needs from maintenance (U.S. cents/ Country adopted date Adopted covered creation board? majority? (number) charges (%)c levy (%) all sources (%) needs (%)d litre petrol) Benin yes 1993 yes 1996 1996 yes no 12 24 52 100 59 5 5 Burundi no -- no -- 2001 yes yes 14 85 60 28 40 4 4 Cameroon yes 1996 nof -- 1998 yes no 13 99 94 60 37 8 10 Central African Republic yes 1990 yes 1990­2005 2000 yes yes 21 91 90 75 20 10 10 Chad yes 1999 yes 1999 2000 yes yes 13 100 59 60 72 26 15 Congo, Rep. nof -- nof -- 2004 yes no 20 -- -- 60 95 -- -- Côte d'lvoire yes 1998 no -- 2001 yes no 15 25 100 80 80 8 2 Djibouti no -- yes 2006 1999 yes no 76 95 0 70 20 -- -- Ethiopia yes 1997 yes 1997 1997 yes no 35 7 99 100 65 9 7 Gabon yes 1997 no -- 1997 yes no 15 0 0 100 70 -- -- Ghana yes 1996 yes 2001 1985 yes yes 12 97 87 100 70 6 6 Guinea yes 1999 yes 2003 2000 yes no 21 54 100 41 67 2 2 Kenya no -- no -- 1999 yes yes 48 99 96 50 55 8 8 Lesotho yes 1995 yes 1995 1995 yes no 10 92 67 50 35 3 3 Madagascar nof -- no -- 1997 yes no 23 76 100 80 30 4 3 Malawi yes 2001 nof -- 1997 no no -- 100 90 40 50 8 6 Mali yes 1998 yes 2002 2000 yes yes 11 33 75 67 40 3 3 Mozambique yes 2002 yes 2002 2003 yes yes 23 93 87 100 60 12 1 Namibia yes 1995 yes 2003 1999 yes yes 16 75 96 80 65 16 1 Niger yes 2004 yes 2005 1999 yes yes 6 96 92 88 35 6 6 Rwanda yes 2005 yes 2005 1998 yes yes 10 65 67 21 25 4 4 Tanzania yes 2003 yes 2001 1998 yes yes 10 89 94 29 76 8 8 Togo yes 1996 no -- 1997 yes yes 10 100 100 56 56 7 7 Zambia yes 2000 yes 2003 1994 yes no 11 16 100 100 30 5 5 Zimbabwe yes 2005 no -- 2001 yes no 15 98 98 30 70 3 3 Average 19 71 79 57 53 7 6 % of direct Road network Allocation of RMF resources (%) Road agency conditionsb channeling Audits of road user RMF Date of Has a Private % in % in Country charges Main roads Rural roads Urban roads overheads Established creation board? majority? good bad Technical Financial Benin 10 96 13 0.3 2 no -- -- -- 41 7 yes yes Burundi 100 70 30 0 5.7 no -- -- -- 10 60 yes yes Cameroon 75 65 12 10 2.5 no -- -- -- 23 35 yes yes Central African Republic 100 75 -- 5 10 no -- -- -- -- -- no no Chad 100 82 0 15 3 no -- -- -- 48 -- yes yes Congo, Rep. 0 20 15 10 5 no -- -- -- -- -- yes yes Côte d'lvoire 25 90 0 10 0 yes 2001 yes no 20 50 no yes Djibouti 0 50 0 20 30 nof -- -- -- 30 10 no yes Ethiopia 100 65 25 10 0 yes 1997 yes no 37 35 yes yes Gabon 0 62 -- 30 2 no -- -- -- -- -- yes yes Ghana 100 37 30 25 1.5 yes 1974 yes no 34 40 yes yes Guinea 59 49 24 22 9 no -- -- -- 22 38 yes yes Kenya 95 57 28 10 3 no -- -- -- 18 33 yes yes Lesotho 93 -- -- -- 3.5 no -- -- -- 20 20 yes yes 435 (continued) 436 % of direct Road network Allocation of RMF resources (%) Road agency conditionsb channeling Audits of road user RMF Date of Has a Private % in % in Country charges Main roads Rural roads Urban roads overheads Established creation board? majority? good bad Technical Financial Madagascar 100 72 14 12 2 nof -- -- -- 25 25 yes yes Malawi 100 45 25 10 5 yes 1998 yes yes 30 36 yes yes Mali 0 70 20 10 2.5 yes 2004 yes no 50 20 yes yes Mozambique 13 35 25 10 1 yes 2003 yes yes 55 28 yes yes Namibia 100 55 33 4.5 1.5 yes 1999 yes yes 13 12 yes yes Niger 0 64 -- 12 4.7 no -- -- -- 30 28 yes yes Rwanda 100 26 10 61 3 nof -- -- -- 20 40 no yes Tanzania 100 70 30 -- -- yes 1999 yes yes 33 31 yes yes Togo 100 74 12 10 4 no -- -- -- 5 32 yes yes Zambia 0 50 25 25 -- yes 2004 yes no 57 22 yes -- Zimbabwe 100 28 28 14 6 yes 2002 yes no 50 45 yes -- Average 63 59 18 15 5 -- -- -- -- 31 31 -- -- Source: SSATP 2007. Note: RMF = road maintenance fund; -- = not available or not relevant. a. Transport policy document may be for roads alone or may be broader. b. Denotes public road networks as a whole. c. Road user charges = fuel levy + road tolls + transit + overloading fees. d. Total maintenance denotes all maintenance works (routine and periodic) excluding rehabilitation. e. This is the fuel levy actually collected (not amount legislated). f. In preparation. Appendix 2l Selected Standards by Network Type, Surface Class, and Traffic Level for 40 Countries (percent) Paved roads Paved roads Unpaved roads Paved roads Unpaved roads Unpaved roads in the primary in the primary in the primary in the secondary in the secondary in the tertiary network with network with network with network with network with network with 10,000 AADT 300 AADT more than 300 AADT more than 300 AADT Country or more or less 300 AADT or less 300 AADT or less Angola 0 42 0 100 0 100 Benin 1 10 34 0 2 95 Botswana 0 23 0 72 60 100 Burkina Faso 0 26 1 45 1 74 Burundi 0 47 0 0 0 100 Cameroon 0 16 36 49 18 94 Central African Republic 0 100 0 0 0 100 Chad 0 25 0 0 1 100 Congo, Dem. Rep. 0 86 0 0 0 100 Congo, Rep. 0 0 0 0 0 99 Côte d'Ivoire 0 24 3 30 7 93 Eritrea 0 36 0 100 0 100 Ethiopia 0 32 40 0 6 74 Gabon 0 10 0 52 0 99 Gambia, The 0 65 0 0 0 100 Ghana 0 13 0 26 0 72 Guinea 1 27 8 80 9 100 Kenya 0 26 3 37 4 89 Lesotho 10 44 18 27 18 65 437 Liberia 0 32 0 100 0 100 Madagascar 1 37 0 60 0 97 (continued) 438 Paved roads Paved roads Unpaved roads Paved roads Unpaved roads Unpaved roads in the primary in the primary in the primary in the secondary in the secondary in the tertiary network with network with network with network with network with network with 10,000 AADT 300 AADT more than 300 AADT more than 300 AADT Country or more or less 300 AADT or less 300 AADT or less Malawi 0 22 13 87 0 81 Mali 0 48 0 59 3 100 Mauritania 0 82 15 100 0 100 Mauritius 16 0 0 0 0 0 Mozambique 0 34 7 72 2 40 Namibia 2 27 0 73 1 82 Niger 0 53 24 0 0 73 Nigeria 3 5 27 23 20 80 Rwanda 0 31 0 0 6 100 Senegal 0 9 0 34 0 77 Sierra Leone 0 69 8 73 1 100 South Africa 20 11 0 100 0 100 Sudan 0 19 2 100 0 100 Swaziland 1 5 62 4 24 100 Tanzania 2 10 27 25 10 100 Togo 6 19 76 0 44 99 Uganda 4 12 26 0 0 100 Zambia 2 41 0 86 6 80 Zimbabwe 0 16 0 90 5 100 Average 2 31 11 43 6 89 Median 0 26 0 36 1 100 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. Note: AADT = annual average daily traffic. The thresholds used in these calculations broadly represent current engineering judgment on the minimum traffic levels necessary to progress from one standard to a higher one. Appendix 2m Preservation Requirements for Securing the Custom Standard over a 20-Year Period for 40 Countries Total Grand total Distribution by network type of grand total (%) Total including including Classified classified unclassified urban Country (US$ millions) (US$ millions) (US$ millions) Primary Secondary Tertiary Unclassified Urban Angola 2,597 2,630 3,345 61 14 3 1 21 Benin 647 663 779 75 5 3 2 15 Botswana 1,613 1,625 1,932 45 31 7 1 16 Burkina Faso 1,044 1,059 1,181 70 13 5 1 10 Burundi 338 350 389 74 11 2 3 10 Cameroon 1,369 1,390 1,763 54 16 8 1 21 Central African Republic 783 799 1,028 41 31 4 2 22 Chad 725 737 944 51 17 9 1 22 Congo, Dem. Rep. 2,973 2,984 3,575 59 18 6 0 17 Congo, Rep. 647 663 779 75 5 3 2 15 Côte d'lvoire 1,694 1,695 2,076 57 21 4 0 18 Eritrea 403 408 485 62 19 2 1 16 Ethiopia 1,817 1,871 1,970 51 20 22 3 5 Gabon 395 402 644 47 10 4 1 38 Gambia, The 296 298 353 62 19 2 0 16 Ghana 1,866 1,891 2,835 22 30 15 1 33 Guinea 908 924 1,056 64 17 5 1 13 Kenya 2,950 2,963 3,238 33 42 16 0 8 439 Lesotho 494 495 540 54 29 8 0 8 (continued) Total Grand total Distribution by network type of grand total (%) Total including including Classified 440 classified unclassified urban Country (US$ millions) (US$ millions) (US$ millions) Primary Secondary Tertiary Unclassified Urban Liberia 367 383 546 45 16 7 3 30 Madagascar 1,459 1,468 1,631 33 42 14 1 10 Malawi 954 959 1,069 52 34 4 0 10 Mali 1,666 1,693 1,839 54 30 7 1 8 Mauritania 588 593 774 68 6 2 1 23 Mauritius 330 331 438 47 22 6 0 24 Mozambique 1,721 1,794 2,009 50 18 18 4 11 Namibia 2,066 2,105 2,448 36 35 14 2 14 Niger 1,026 1,034 1,095 71 11 11 1 6 Nigeria 11,305 11,462 12,916 55 30 3 1 11 Rwanda 271 291 420 53 11 1 5 31 Senegal 1,185 1,190 1,313 59 21 10 0 9 Sierra Leone 503 504 704 46 21 5 0 28 South Africa 14,696 15,044 17,802 39 29 14 2 15 Sudan 1,471 1,483 1,925 54 20 3 1 23 Swaziland 414 415 441 62 30 2 0 6 Tanzania 2,701 2,714 3,130 34 49 4 0 13 Togo 741 747 820 75 14 1 1 9 Uganda 2,710 2,721 2,986 41 44 7 0 9 Zambia 1,710 1,751 1,899 43 31 16 2 8 Zimbabwe 2,409 2,450 2,773 51 31 5 1 12 Total 73,851 74,976 87,888 Average 1,846 1,874 2,197 53 23 7 1 16 Median 1,114 1,124 1,247 53 20 5 1 14 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. Note: The "custom" standard applies the same required standard to each road network type, with primary roads maintained to good standard and secondary only to fair standard. For the "optimal" standard, RONET evaluates road classes defined in terms of both engineering characteristics and traffic volumes and identifies for each road class the standard that maximizes the society benefits Net Present Value, at a given discount rate. Appendix 2n Preservation Requirements for Securing the Optimal Standard over a 20-Year Period for 40 Countries Total Grand total Distribution by network type of grand total (%) Total including including Classified classified unclassified urban Country (US$ millions) (US$ millions) (US$ millions) Primary Secondary Tertiary Unclassified Urban Angola 1,208 1,276 1,609 63 8 4 4 21 Benin 525 558 612 78 5 3 5 9 Botswana 894 919 1,062 34 42 9 2 13 Burkina Faso 656 688 745 60 16 12 4 8 Burundi 162 187 206 65 11 3 12 9 Cameroon 1,105 1,149 1,323 59 17 8 3 13 Central African Republic 495 529 636 38 34 5 5 17 Chad 468 491 588 55 13 12 4 16 Congo, Dem. Rep. 1,443 1,467 1,742 52 21 10 1 16 Congo, Rep. 525 558 612 78 5 3 5 9 Côte d'lvoire 1,149 1,152 1,329 58 24 5 0 13 Eritrea 261 271 307 65 18 2 3 12 Ethiopia 1,171 1,283 1,329 29 25 34 8 3 Gabon 248 262 375 52 9 6 4 30 Gambia, The 167 170 196 62 20 3 2 13 Ghana 1,498 1,563 2,469 19 21 21 3 37 Guinea 650 682 744 63 18 6 4 8 Kenya 2,264 2,292 2,420 26 42 25 1 5 Lesotho 461 464 485 46 36 13 1 4 Liberia 240 273 349 44 16 8 10 22 441 Madagascar 820 838 914 21 53 16 2 8 (continued) Total Grand total Distribution by network type of grand total (%) 442 Total including including Classified classified unclassified urban Country (US$ millions) (US$ millions) (US$ millions) Primary Secondary Tertiary Unclassified Urban Malawi 682 692 744 42 42 8 1 7 Mali 834 891 959 40 36 11 6 7 Mauritania 179 190 275 54 7 5 4 31 Mauritius 246 248 259 82 11 2 1 4 Mozambique 1,206 1,370 1,470 31 18 33 11 7 Namibia 1,615 1,696 1,856 23 37 26 4 9 Niger 524 540 569 50 16 26 3 5 Nigeria 9,045 9,373 10,050 52 33 4 3 7 Rwanda 143 186 245 44 12 2 17 24 Senegal 873 884 941 63 21 9 1 6 Sierra Leone 344 346 440 47 26 6 0 21 South Africa 8,705 10,166 11,695 40 20 14 12 13 Sudan 1,044 1,069 1,274 60 15 7 2 16 Swaziland 339 341 353 54 40 2 1 3 Tanzania 2,491 2,517 2,711 26 59 7 1 7 Togo 651 664 698 72 20 1 2 5 Uganda 2,113 2,135 2,259 50 29 14 1 5 Zambia 1,178 1,349 1,418 33 27 23 12 5 Zimbabwe 1,512 1,598 1,748 51 27 8 5 9 Total 50,136 53,331 60,016 Average 1,253 1,333 1,500 50 24 10 4 12 Median 751 765 829 51 20 8 3 9 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. Note: The "custom" standard applies the same required standard to each road network type, with primary roads maintained to good standard and secondary only to fair standard. For the "optimal" standard, RONET evaluates road classes defined in terms of both engineering characteristics and traffic volumes and identifies for each road class the standard that maximizes the society benefits Net Present Value, at a given discount rate. Appendix 2o Custom Standard 20-Year Preservation Needs by Work Type for 40 Countries (US$ millions) Distribution by work type Average annual Average annual Recurrent Country grand total Years 1­5 Years 6­20 Road rehabilitation Periodic maintenance maintenance Angola 167 281 129 83 42 43 Benin 39 78 26 25 5 9 Botswana 97 148 79 50 22 24 Burkina Faso 59 58 59 21 19 18 Burundi 19 28 17 9 5 6 Cameroon 88 187 55 55 12 21 Central African Republic 51 68 46 15 17 19 Chad 47 78 37 18 13 17 Congo, Dem. Rep. 179 312 134 72 55 52 Congo, Rep. 39 78 26 25 5 9 Côte d'lvoire 104 176 80 47 33 24 Eritrea 24 39 19 10 8 7 Ethiopia 98 143 84 20 55 23 Gabon 32 66 21 19 4 9 Gambia, The 18 34 12 10 4 4 Ghana 142 201 122 48 58 36 Guinea 53 108 35 29 11 13 Kenya 162 237 137 57 62 43 Lesotho 27 48 20 11 10 6 Liberia 27 42 22 11 8 9 Madagascar 82 147 60 42 19 21 443 (continued) Distribution by work type 444 Average annual Average annual Recurrent Country grand total Years 1­5 Years 6­20 Road rehabilitation Periodic maintenance maintenance Malawi 53 67 49 19 16 18 Mali 92 124 81 38 25 29 Mauritania 39 56 33 20 7 11 Mauritius 22 22 22 7 9 5 Mozambique 100 146 85 43 33 25 Namibia 122 165 108 45 41 37 Niger 55 73 49 25 16 14 Nigeria 646 1,228 452 274 238 133 Rwanda 21 32 17 11 4 6 Senegal 66 151 37 43 9 13 Sierra Leone 35 63 26 16 9 10 South Africa 890 1,450 703 424 240 226 Sudan 96 159 75 38 30 29 Swaziland 22 29 20 4 12 6 Tanzania 157 195 144 45 61 51 Togo 41 94 23 23 11 7 Uganda 149 262 112 61 45 44 Zambia 95 190 63 55 20 20 Zimbabwe 139 223 111 63 44 32 Total 4,394 7,284 3,431 1,929 1,338 1,127 Average 110 182 86 48 33 28 Median 62 116 52 27 17 19 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. Note: The "custom" standard applies the same required standard to each road network type, with primary roads maintained to good standard and secondary only to fair standard. For the "optimal" standard, RONET evaluates road classes defined in terms of both engineering characteristics and traffic volumes and identifies for each road class the standard that maximizes the society benefits Net Present Value, at a given discount rate. Appendix 2p Optimal Standard 20-Year Preservation Needs by Work Type for 40 Countries (US$ millions) Average annual Average annual Distribution by work type Country grand total Years 1­5 Years 6­20 Road rehabilitation Periodic maintenance Recurrent maintenance Angola 80 145 59 41 22 18 Benin 31 65 19 21 4 6 Botswana 53 79 45 26 16 11 Burkina Faso 37 36 38 11 15 12 Burundi 10 14 9 4 3 3 Cameroon 66 140 42 45 9 12 Central African Republic 32 27 34 4 15 12 Chad 29 37 27 9 10 10 Congo, Dem. Rep. 87 91 86 17 47 23 Congo, Rep. 31 65 19 21 4 6 Côte d'lvoire 66 128 46 36 18 13 Eritrea 15 25 12 6 6 3 Ethiopia 66 80 62 16 33 17 Gabon 19 37 13 12 2 4 Gambia, The 10 15 8 6 2 2 Ghana 123 167 109 40 55 29 Guinea 37 74 25 22 8 7 Kenya 121 151 111 38 51 31 Lesotho 24 41 19 10 10 4 Liberia 17 23 16 6 6 5 Madagascar 46 82 34 25 11 9 Malawi 37 53 32 16 12 9 Mali 48 55 46 14 18 15 445 Mauritania 14 17 13 6 3 5 (continued) 446 Average annual Average annual Distribution by work type Country grand total Years 1­5 Years 6­20 Road rehabilitation Periodic maintenance Recurrent maintenance Mauritius 13 17 12 2 8 3 Mozambique 74 105 63 33 25 16 Namibia 93 133 79 33 35 24 Niger 28 38 25 13 8 7 Nigeria 502 980 343 249 172 82 Rwanda 12 14 12 7 2 3 Senegal 47 115 24 34 6 7 Sierra Leone 22 31 19 10 7 5 South Africa 585 872 489 259 198 128 Sudan 64 104 50 25 23 16 Swaziland 18 20 17 3 11 4 Tanzania 136 175 122 38 66 31 Togo 35 77 21 20 11 4 Uganda 113 170 94 38 50 25 Zambia 71 145 46 45 14 11 Zimbabwe 87 148 67 40 30 17 Total 3,001 4,790 2,404 1,302 1,048 651 Average 75 120 60 33 26 16 Median 41 75 34 20 11 10 Source: Calculations by Alberto Nogales based on AICD RONET summary outputs, June 2010. Note: The "custom" standard applies the same required standard to each road network type, with primary roads maintained to good standard and secondary only to fair standard. For the "optimal" standard, RONET evaluates road classes defined in terms of both engineering characteristics and traffic volumes and identifies for each road class the standard that maximizes the society benefits Net Present Value, at a given discount rate. Roads 447 References Archondo Callao, R. 2009. "RONET--Road Network Evaluation Tools: User Guide, Version 2.0." Sub-Saharan Africa Transport Policy Program Working Paper 89A, World Bank, Washington DC. SSATP (Sub-Saharan Africa Transport Policy Program). 2007. "Road Maintenance Indicators Matrix." World Bank, Washington, DC. World Health Organization. 2009. "Global Status Report on Road Safety: Time for Action." World Health Organization, Geneva. APPENDIX 3 Rail Transport 449 Appendix 3a Rail Networks in Africa 450 Lines (kms) Network density Country Railway operator Total Operating km/1,000 km2 km/ million pop. Gaugea Angola Amboin 123 0 -- -- 3 CFB 1,348 246 2.3 222 1 CFL 510 181 -- -- 1 CFM-Angola 907 425 -- -- 1 Benin OCBN 579 438 5.1 66 2 Botswana BRC 882 882 1.5 443 1 Burkina Faso/Côte d'Ivoire Sitarail 1,250 1,250 2.0 35 2 Cameroon CAMRAIL 1,100 1,016 2.3 58 2 Congo, Dem. Rep. CFMK 366 366 1.9 -- 1 CFU 1,028 0 -- -- 3 SNCC 3,641 2,200 2.1 73 1 Congo, Rep. CFCO 795 610 2.3 198 1 Djibouti/Ethiopia CDE 781 781 1.0 100 2 Eritrea ERA 117 0 1.0 21 3 Gabon SETRAG 649 649 2.4 428 4 Ghana GRC 947 947 4.0 40 1 Guinea Bauxitelines 383 -- -- -- 4 ONCFG 662 0 4.3 104 2 Kenya KRC 1,894 1,894 3.8 57 2 Magadi 171 -- -- -- 2 Liberia Bong 77 77 -- -- 4 LAMCO 267 0 3.8 124 4 NIOC 84 -- -- -- 1 Madagascar FCE 236 0 -- -- 2 Madarail 650 650 1.5 43 2 Malawi CEAR 797 710 6.7 56 1 Mali RNCFM 57 0 -- -- -- Mali/Senegal Transrail 1,229 1,229 5.4 77 2 Mauritania SNIM 704 -- 0.7 225 4 Mozambique CCFB 987 317 3.9 144 1 CDN 872 611 -- -- 1 CFM-Mozambique 1,269 1,129 -- -- 1 Namibia TransNamib 2,382 1,683 2.9 1,129 1 Nigeria Central Railways 52 0 -- -- 4 NRC 3,505 3,505 3.8 23 1 Senegal Sefincs -- -- -- -- 2 SNCS 408 40 -- -- 2 TrainBleu -- -- -- -- 2 Sierra Leone MMR 84 0 1.2 13 1 South Africa Coalex 574 574 -- -- 1 Metrorail 1,318 1,318 -- -- 1 Orex 880 880 -- -- 1 Spoornet 18,793 18,793 17.7 440 1 Sudan SR-Sudan 4,680 2,252 -- -- 1 Swaziland SR-Swaziland 301 301 17.3 268 1 Tanzania TRC 2,722 2,722 3.8 87 2 Tanzania/Zambia TAZARA 1,860 1,860 -- -- 1 Togo CFTB 522 -- 9.2 87 2 Uganda URC 1,244 261 5.3 38 2 Zambia RSZ 1,273 1,273 3.2 203 1 Zimbabwe BBR 150 150 -- -- 1 NRZ 3,077 2,759 8.3 283 1 Source: Bullock 2009. 451 Note: -- = not available. For full names of railway companies, see list following appendix 3e. a. 1 = Cape; 2 = Metre; 3 = Narrow; 4 = Standard. 452 Appendix 3b Production Structure of African Railways, Average 1995­2005 Transport task (million units) Proportion of total task (%) Average haul (km) Country Railway operator Passenger-km Net tonne-km Passenger Freight Passenger Freight Angola CFM-Angola 89 472 16 84 91 146 Benin OCBN 18 24 43 57 153 462 Botswana BRC 94 631 13 87 237 342 Burkina Faso/Côte d'Ivoire Sitarail 25 670 0 100 140 899 Cameroon CAMRAIL 308 1,119 22 78 292 639 Congo, Dem. Rep. CFMK 3 57 5 95 91 345 SNCC 70 444 14 86 195 382 Congo, Rep. CFCO 167 264 39 61 266 379 Djibouti/Ethiopia CDE 82 97 0 100 253 404 Gabon SETRAG 87 2,208 4 96 399 563 Ghana GRC 64 224 22 78 30 123 Kenya KRC 380 1,553 20 80 780 783 Madagascar FCE 7 1 88 13 65 111 Madarail 1 84 1 99 106 341 Malawi CEAR 20 39 34 66 44 163 Mali RNCFM 196 271 42 58 356 709 Mali/Senegal Transrail 113 409 22 78 991 965 Mozambique CCFB 3 172 2 98 13 277 CDN 80 123 39 61 143 564 Namibia TransNamib 45 1,247 3 97 450 636 Nigeria NRC 174 77 69 31 73 819 Senegal SNCS 173 485 26 74 34 168 South Africa Spoornet 991 109,721 1 99 320 602 Sudan SR-Sudan 40 766 3 97 625 606 Swaziland SR-Swaziland 0 680 0 100 -- 170 Tanzania TRC 421 1,107 28 72 696 916 Tanzania/Zambia TAZARA 241 937 20 80 219 1,483 Togo CFTB -- 150 -- -- -- -- Uganda URC 0 186 0 100 152 219 Zambia RSZ 69 433 14 86 358 382 Zimbabwe NRZ 457 1,404 25 75 250 360 Source: Bullock 2009. Note: -- = not available. For full names of railway companies, see list following appendix 3e. 453 Appendix 3c Rail Passenger Traffic 454 Traffic unit, Traffic density, Passenger passengers passengers Trip length per (number / Passengers as (million (1,000 passenger- passenger (km/ Railway year) % of total passenger-km) km/route-km) passenger) Country operator 2000 2005 transport task 2000 2005 2005 2000 2005 Angola CFM-Angola 1,469 975 16 52 89 188 35 91 Benin OCBN 577 118 43 100 18 119 173 153 Botswana BRC 310 396 13 75 94 118 242 237 Burkina Faso/Côte d'Ivoire Sitarail 300 0 0 126 -- 46 420 -- Cameroon CAMRAIL 1,266 1,053 22 384 308 318 303 292 Congo, Dem. Rep. CFMK 155 33 5 16 3 11 103 91 SNCC 1,307 359 14 188 70 60 144 195 Congo, Rep. CFCO 546 628 39 84 167 195 154 266 Djibouti/Ethiopia CDE 710 -- 0 145 -- 160 204 -- Gabon SETRAG 237 218 4 95 87 139 401 399 Ghana GRC 844 2,134 22 83 64 73 98 30 Kenya KRC 422 487 20 311 380 176 737 780 Madagascar Madarail -- 11 1 -- 1 0 -- -- Malawi CEAR 381 452 34 25 20 38 66 44 Mali RNCFM 682 0 42 204 -- -- 299 -- Mali/Senegal Transrail -- 114 22 -- 113 92 -- 991 Mozambique CCFB 540 229 2 18 3 44 33 13 CDN 886 561 39 60 80 103 68 143 Namibia TransNamib 154 -- 3 66 45 33 -- -- Nigeria NRC 2,610 2,399 69 170 174 31 65 73 Senegal SNCS 2,783 0 26 143 -- -- 51 -- South Africa Spoornet 4,216 3,100 1 1,370 991 60 -- 320 Sudan SR-Sudan 410 64 3 129 40 26 315 625 Tanzania TRC 631 605 28 428 421 163 678 696 Tanzania/Zambia TAZARA 1,543 1,100 20 517 241 148 335 219 Uganda URC 0 0 0 0 0 0 -- -- Zambia RSZ 623 193 14 147 69 92 236 358 Zimbabwe NRZ 1,691 -- 25 565 -- 166 -- -- Low-income, fragile 196 20 -- 37 124 -- 186 -- Low-income, nonfragile 415 33 -- 122 158 -- 231 -- Middle-income 1,165 15 -- 282 1,962 -- 310 -- Resource-rich 790 29 -- 133 151 -- 169 -- State-owned railways 862 30 -- 202 780 -- 200 -- Non-state-owned railways 310 22 -- 108 95 -- 315 -- Africa 586 27 -- 145 569 -- 226 -- Source: Bullock 2009. Note: -- = not available. For full names of railway companies, see list following appendix 3e. Passengers as percentage of total transport task is shown for 2005, where available; otherwise, for 2000. 455 456 Appendix 3d Pricing and Institutions Concessioned company Freight average (1 = concessioned, Private company yield, 2000­05 Passenger average Railway 0 = nonconcessioned) (1 = private, 0 = nonprivate) (U.S. cents/net yield, 2000­05 Country operator tonne-km) (U.S. cents/passenger-km) 2005 2005 Angola CFM-Angola -- -- 0 0 Benin OCBN 5.8 2.0 0 0 Botswana BRC 2.4 1.3 0 0 Burkina Faso/Côte d'Ivoire Sitarail 5.5 3.3 1 0 Cameroon CAMRAIL 5.2 2.2 1 0 Congo, Dem. Rep. CFMK 13.7 4.2 0 0 SNCC 12.5 3.1 1 0 Congo, Rep. CFCO 10.7 5.6 1 0 Djibouti/Ethiopia CDE 12.5 3.1 1 0 Gabon SETRAG 2.5 8.6 1 0 Ghana GRC 4.4 2.4 1 0 Kenya KRC 3.8 0.6 1 0 Madagascar Madarail 5.1 -- 1 0 Malawi CEAR 5.8 1.0 1 0 Mali RNCFM -- -- 0 0 Mali/Senegal Transrail 3.3 2.2 1 0 Mozambique CCFB 3.3 0.5 1 0 CDN 5.4 0.9 1 0 Namibia TransNamib -- -- 0 0 Nigeria NRC -- -- 1 0 Senegal SNCS -- -- 0 0 South Africa Spoornet -- -- 0 0 Sudan SR-Sudan -- -- 0 0 Tanzania TRC 4 1.6 0 0 Tanzania/Zambia TAZARA 3 1.1 0 0 Uganda URC 15.2 2.3 1 0 Zambia RSZ 3.9 0.8 1 0 Zimbabwe NRZ -- -- 0 0 Low-income, fragile 11 4 -- -- Low-income, nonfragile 5 1 -- -- Middle-income 7 2 -- -- Resource-rich 6 4 -- -- State-owned railways 7 3 -- -- Non-state-owned railways 5 2 -- -- Africa 6 2 -- -- Source: Bullock 2009. Note: -- = not available. For full names of railway companies, see list following appendix 3e. 457 458 Appendix 3e Factor Productivity Locomotive Carriage productivity Wagon Labor productivity productivity (1,000 (million traffic productivity (1,000 (1,000 traffic units passenger-km per units per net tonne-km per Railway Employees (no.) per employee) carriage) locomotive) wagon) Country operator 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 Angola CFM-Angola 6,492 967 43 580 1,793 4,045 10 30 263 950 Benin OCBN 1,204 1,040 157 40 5,263 900 13 3 273 74 Botswana BRC 1,177 1,004 945 722 -- -- -- -- -- -- Burkina Faso/Côte d'Ivoire Sitarail 1,712 1,393 379 481 4,500 -- 32 35 682 1,020 Cameroon CAMRAIL 2,711 2,367 477 603 5,053 4,738 18 26 783 868 Congo, Dem. Rep. CFMK 2,945 1,950 57 221 1,527 3,212 14 27 104 300 SNCC 3,401 3,359 29 18 291 64 16 10 215 257 Djibouti/Ethiopia CDE 15,266 13,619 41 38 764 275 7 4 302 317 Gabon SETRAG 1,400 1,291 1,284 1,778 2,065 1,891 30 39 749 902 Ghana GRC 4,376 3,425 57 84 539 416 6 7 320 458 Kenya KRC 9,700 9,500 170 203 948 1,159 20 24 187 218 Madagascar Madarail -- -- 0 0 -- -- -- -- -- 449 Malawi CEAR 642 -- 165 0 1,389 1,176 6 -- 186 -- Mali RNCFM 1,475 -- 269 0 5,368 -- 17 -- 745 -- Mali/Senegal Transrail -- 1,540 0 339 -- -- -- 40 -- 804 Mozambique CCFB 5,859 623 42 281 818 750 14 13 263 476 CDN 2,692 286 78 710 3,158 3,333 12 25 560 260 Namibia TransNamib 2,213 -- 440 0 -- -- 21 -- 568 -- Nigeria NRC 14,507 6,748 17 37 566 737 8 13 62 59 Senegal SNCS 1,490 -- 455 0 1,324 -- 15 -- 713 -- South Africa Spoornet 40,808 33,467 2,646 3,308 649 -- 33 -- 921 -- Sudan SR-Sudan 14,021 -- 96 0 772 -- 12 -- 282 -- Tanzania TRC 8,947 5,900 195 259 3,035 3,264 23 25 968 583 Tanzania/Zambia TAZARA 4,652 3,346 279 352 5,744 2,770 15 13 655 551 Uganda URC 1,566 1,193 134 156 -- -- 8 9 165 144 Zambia RSZ 3,400 1,000 205 502 1,986 3,286 11 25 95 377 Zimbabwe NRZ 9,536 -- 408 0 -- -- -- 8 -- -- Low-income, fragile -- 13,619 -- 38 -- 257 -- 7 -- -- Low-income, nonfragile -- 4,387 -- 240 -- 1,610 -- 19 -- -- Middle-income -- 32,443 -- 3,258 -- -- -- 136 -- -- Resource-rich -- 1,428 -- 982 -- 2,518 -- 32 -- 810 State-owned railways -- 31,195 -- 3,118 -- 1,664 -- 25 -- 708 Non-state-owned railways -- 3,530 -- 387 -- 2,945 -- 24 -- 510 Africa -- 2,9578 -- 2,998 -- 1,994 -- 38 -- 632 Source: Bullock 2009. Note: -- = not available. For full names of railway companies, see list following appendix 3e. 459 460 Africa's Transport Infrastructure Railway Names Railway Country acronym Railway full name Angola CFB Caminho de Ferro de Benguela Angola CFL Caminho de Ferro de Luanda Angola CFM-Angola Caminho de Ferro de Moçâmedes Benin OCBN Organisation Commune Benin-Niger des Chemins de fer et des Transports Botswana BRC Botswana Railways Burkina Faso/ Sitarail Société Internationale de Transport Africain par Rail Côte d'Ivoire Cameroon CAMRAIL Cameroon National Railway Congo, SNCC Société des Chemins de Fer du Congo Dem. Rep. CFMK Chemins de Fer Matadi Kinshasa CFU Chemin de Fer Urbain Congo, Rep. CFCO Chemin de Fer du Congo Océan Djibouti/Ethiopia CDE Chemin de Fer Djibouti-Ethiopien Eritrea ERA Eritrean Railway Gabon SETRAG Société d' Exploitation du Transgabonais Ghana GRC Ghana Railway Company Guinea ONCFG Office National des Chemins de Fer de Guinee Kenya KRC Kenya Railways Corporation Liberia LAMCO Liberian-American-Swedish Minerals Company Liberia NIOC National Iron Ore Company of Liberia Madagascar FCE Fianarantsoa Côte Est Madagascar Madarail Private company Malawi CEAR Central East African Railways Mali RNCFM Réseau National des Chemins de Fer Malgache Mauritania SNIM Société Nationale Industrielle et Minière de Mauritanie Mozambique CCFB Campanhia dos Caminhos de Ferro da Beira Mozambique CDN Corredor de Desenvolvimento do Norte Mozambique CFM-Mozambique Portos e Caminhos de Ferro de Moçambique Nigeria NRC Nigerian Railway Corporation Senegal SNCS Société Nationale de Chemins de Fer du Senegal Sierra Leone MMR Expected to be reopened soon Sudan SR-Sudan Sudan Railways Corporation Swaziland SR-Swaziland Swaziland Railway (continued) Rail Transport 461 Railway Country acronym Railway full name Tanzania TRC Tanzania Railways Corporation Tanzania/Zambia TAZARA Tanzania-Zambia Railway Authority Togo CFTB Chemin de Fer du Togo Uganda URC Uganda Railways Corporation Zambia RSZ Railway Systems of Zambia Limited Zimbabwe BBR Beitbridge Bulawayo Railway Zimbabwe NRZ National Railways of Zimbabwe Reference Bullock, R. 2009. "Railways in Sub-Saharan Africa." Africa Infrastructure Country Diagnostic Background Paper 17, World Bank, Washington, DC. APPENDIX 4 Airports and Air Transport Appendix 4a Data Sources for Air Transport Analysis Air traffic analysis is highly data intensive. Unfortunately, because of limitations in both budget and capacities, those countries most in need of development aid are also those that have the most difficulties collecting and reporting vital data. The standard source for traffic data collected by airlines or airports is the International Civil Aviation Organization (ICAO). Yet passenger counts from African airlines are often kept on paper ledgers with no com- puterization and, in many cases, are never submitted to ICAO. For many African countries, the gaps in reporting can be greater than five years; therefore, alternative sources of data must be tapped. An excellent approximation of air traffic is the capacity offered. If one assumes that no airline would, over time, fly an aircraft with too few passengers to make the flight economically feasible, at any given point, 50 percent to 70 percent of the seat capacity offered on a route closely approximates the actual traffic. In addition, even with changes in load fac- tor, the overall time trend of seat capacity approximates traffic trends. Hence, data published by airlines in reservation systems could substitute for travel data. Such data are readily available, are highly granular, and provide a wealth of information not only on the seat capacity, but also on 463 464 Africa's Transport Infrastructure the type of aircraft, the frequency of the routes, and the scheduled times of the flight. Today, there are two main sources of these data: the Official Airline Guide and Seabury's Airline Data Group (ADG). Both sources depend on airlines to report their routes, and both have captured 99 percent of scheduled airline data, with about 900 to 1,000 airlines participating. Though the Official Airline Guide is the more established collector, both companies enjoy excellent industry reputations and are endorsed by the International Air Transport Association. For the studies on Africa undertaken by the World Bank, ADG's data were used. A total of 12 snapshots in time was assembled, 4 each for the years 2001, 2004, and 2007. In order to ensure the capture of seasonal trends, the four samples for each year consisted of data for one week in the months of February, May, August, and November. To annu- alize these figures, the total of the four observations for one year was multiplied by 13. The data consist of one record of each flight occurring during the sampled week, with relevant entries for the origin and destination air- ports, the changeover airport in the case of one-stop flights, the number of kilometers for the flight, the duration of the flight, the number of seats available on the flight, the number of times the flight occurred dur- ing the week, which weekdays the flight was scheduled, the aircraft type, the marketing operator as well as the actual operator, and various other flags of potential analytical utility. Using Microsoft Access, the data were normalized and linked to other relevant tables (some of which were from other sources) to develop a relational database for summarization and querying. One important additional adjustment was made: flights going from one airport to another final destination with one or more stops in between had their capacity allocated evenly between each leg. This implies that a flight from airport A to airport C via airport B would have only half the capac- ity going from airport A to C, while the other half would deplane at air- port B. This methodology prevents double counting of capacity for multilegged flights. For assurance that the approximations used were reasonably accurate, some of the airport aggregates were compared to actual data when avail- able from ICAO. The ratio of seats versus reported traffic hints at a load factor of about 65 percent to 69 percent for those routes tested--a sound figure supporting the credibility of the data. Airports and Air Transport 465 The data are particularly helpful in capturing trends in city and coun- try pairs, fleet renewal (in most cases the type of aircraft is provided, down to the series number, such as Boeing 737-100 versus 737-800), and airline market share. Nevertheless, the data reflect only scheduled and advertised services. An airline with no reservation system that issues paper tickets at the airport and provides a schedule only via a chalkboard or a printed flyer would not be captured. For example, the ADG data show virtually no older, Eastern-bloc-built aircraft operating in Africa, yet anecdotal evidence and accident statistics offer evidence of such operations. The overall portion of this market is suspected to be rela- tively small, though it has a high profile where incidents and accidents are concerned. Since central data collection in Africa is still in a developmental stage, data had to be drawn from diverse sources. A questionnaire with exten- sive details on such things as civil aviation budgets, airport charges, and the number of employees within the civil aviation authority was sent to all 54 African countries. Twenty countries returned the questionnaires, with various levels of completion as their resources allowed. When a true comparative sample set was derived from the questionnaires, it has been applied in this report. However, since the questionnaire was large, and many sections were not completed by the civil aviation authorities, the sample size per answer was often very small. In terms of air navigation and air traffic control infrastructure, reports provided by the Air Navigation Bureau of ICAO provided the most com- prehensive inventory. Spot checks with data returned from the question- naires agreed with data from the Air Navigation Bureau. Data regarding airport infrastructure were gleaned from various sources. Satellite images from a commonly available satellite image serv- ice provided information regarding overall airport and runway condition. All airports receiving scheduled services were surveyed, and roughly 66 percent had images of sufficient quality to draw conclusions. Expert, on-the-ground observational inputs for a sample of 23 airports confirmed the general conclusions drawn from the satellite images. Additional infor- mation for each airport was researched using common data sources, including Jeppesen. Since ICAO does not keep a central database regarding airport termi- nal capacity, data collected by http://www.azworldairports.com, a pub- lisher in the United Kingdom that provides self-reported information from the largest African airports, were used. 466 Africa's Transport Infrastructure Appendix 4b Airports Scheduled Airports with passenger scheduled, advertised carriers based in Total airports service (number) country (number)a (number)a Country 2001 2004 2007 2007 2007 Angola 2 11 13 2 24 Benin 1 1 1 7 7 Botswana 4 4 4 1 101 Burkina Faso 2 2 2 2 25 Burundi 1 1 1 1 3 Cameroon 9 6 3 3 38 Cape Verde 7 7 7 2 7 Central African Republic 1 1 1 0 38 Chad 1 1 1 1 40 Comoros 2 4 5 2 4 Congo, Dem. Rep. 8 8 8 53 239 Congo, Rep. 2 6 2 2 33 Côte d'Ivoire 1 1 1 3 25 Equatorial Guinea 1 1 2 5 2 Eritrea 1 1 1 1 4 Ethiopia 29 26 6 1 41 Gabon 12 3 8 5 30 Gambia, The 1 1 1 4 1 Ghana 1 1 4 2 8 Guinea 1 1 1 0 14 Guinea-Bissau 1 1 1 2 1 Kenya 13 12 10 9 172 Lesotho -- 1 1 0 10 Liberia 1 2 2 1 10 Madagascar 40 36 12 1 56 Malawi 5 3 3 1 23 Mali 6 2 1 5 26 Mauritania 8 8 1 0 20 Mauritius 2 2 2 1 2 Mozambique 8 12 12 1 21 Namibia 11 8 7 1 25 Niger 2 2 1 1 18 Nigeria 4 5 14 7 46 Rwanda 1 1 1 2 7 São Tomé and Príncipe 1 1 1 1 2 Senegal 3 4 4 2 13 Seychelles 1 2 2 1 14 (continued) Airports and Air Transport 467 Scheduled Airports with passenger scheduled, advertised carriers based in Total airports service (number) country (number)a (number)a Country 2001 2004 2007 2007 2007 Sierra Leone 1 1 1 8 8 Somalia 5 3 7 0 14 South Africa 10 21 20 13 195 Sudan 12 11 11 5 44 Swaziland 1 1 1 5 11 Tanzania 16 11 12 3 72 Togo 1 1 1 1 7 Uganda 6 2 4 1 12 Zambia 8 6 6 1 69 Zimbabwe 5 3 4 3 129 Low-income, fragile 9.6 8.5 5.6 2.6 34.3 Low-income, nonfragile 2.1 2.0 2.4 5.3 33.3 Middle-income 8.1 7.9 7.2 3.3 36.9 Resource-rich 5.7 5.6 6.7 3.4 36.2 Africa 6.4 6.0 5.3 3.7 35.0 Source: Africa Infrastructure Country Diagnostic (AICD) data. Note: -- = not available. a. Data are not available for 2001, 2004. Appendix 4c City Pairs Served Domestic (number) International (number) Country 2001 2004 2007 2001 2004 2007 Angola 12 17 21 20 17 21 Benin -- -- -- 20 14 20 Botswana 4 6 4 7 9 8 Burkina Faso 1 1 1 12 14 13 Burundi -- -- -- 5 3 5 Cameroon 17 10 3 33 30 25 Cape Verde 14 11 10 20 16 29 Central African Republic -- -- -- 11 5 3 Chad -- -- -- 17 9 8 Comoros 1 4 7 18 24 19 Congo, Dem. Rep. 13 17 25 24 19 24 Congo, Rep. 4 8 1 31 22 21 (continued) 468 Africa's Transport Infrastructure Domestic (number) International (number) Country 2001 2004 2007 2001 2004 2007 Côte d'Ivoire -- -- -- 42 30 29 Equatorial Guinea -- -- 1 13 9 18 Eritrea 1 -- 3 10 10 14 Ethiopia 76 50 45 46 47 52 Gabon 18 11 9 25 24 13 Gambia, The -- -- -- 12 10 14 Ghana -- -- 5 35 26 30 Guinea -- -- -- 14 9 10 Guinea-Bissau -- -- -- 9 2 3 Kenya 25 18 15 60 66 77 Lesotho -- -- -- 1 1 1 Liberia -- -- -- 9 8 11 Madagascar 133 97 63 16 21 32 Malawi 9 8 3 15 20 12 Mali 15 1 -- 24 19 18 Mauritania 8 10 1 15 10 10 Mauritius 1 1 1 32 32 34 Mozambique 22 25 30 13 22 32 Namibia 22 13 8 15 13 14 Niger -- -- -- 18 14 9 Nigeria 17 6 22 50 41 51 Rwanda 3 -- -- 10 6 7 São Tomé and Príncipe 1 -- -- 5 3 5 Senegal 4 4 4 33 34 43 Seychelles 1 1 1 16 16 12 Sierra Leone -- -- -- 8 9 11 Somalia 9 6 9 14 13 19 South Africa 48 48 36 101 100 115 Sudan 22 18 13 33 35 32 Swaziland -- -- -- 4 4 4 Tanzania 44 21 19 41 44 38 Togo -- -- -- 18 14 12 Uganda 8 1 4 21 10 19 Zambia 18 9 9 17 16 16 Zimbabwe 7 2 5 31 20 23 Low-income, fragile 28.8 21.2 17.8 30.6 30.0 36.2 Low-income, nonfragile 5.3 7.3 9.8 15.3 11.9 13.5 Middle-income 22.2 18.4 14.7 51.5 50.9 54.8 Resource-rich 15.4 11.3 9.9 26.6 22.6 22.8 Africa 20.4 16.5 14.0 30.7 28.8 32.1 Source: AICD data. Note: -- = not available. Airports and Air Transport 469 Appendix 4d Installation of Ground-Based Navigational Aids in Africa 20° W 10° W 0° 10° E 20° E 30° E 40° E 50° E 40° N 40° N Dellys Zémmouri Béjaïa Béni-Amrane Eth Tabarka Alger Jijel Annaba Tunis Cherchell Cap Bon Tétouan Mostaganem Constantine Monastir Tanger Oran Bou-Saada Tébéssa Ben Aoun Sfax Rabat Al Hoceima Oujda Tiaret El Bayadh El Oued Gafsa Djerba TUNISIA Casablanca Fès Tozeur Zawia Touggourt Errachidia Méchéria Tripoli MOROCCO Marrakech Ghardaïa El-Borma Beni Walid Benghazi El Daba Alexandria Baltim Ouarzazate El Goléa Hassi- La Palma Fuerteventura Agadir Méssaoud Ghadames Cairo Taba 30° N Tenerife Norte Lanzarote 30° N Béni-Abbès Timimoun Fayoum Canary Islands Hierro Nuweibaa (Sp.) Tenerife Sur Gran An-tan Tindouf Bordj Omar Driss Zarzaitine LIBYA Sarir Sainte-Catherine Sharm el Sheikh El Aaiun In Salah Sebha Asyut Hurghada FORMER Canari Réggan Illizi ARAB Luxor SPANISH ALGERIA Djanet Kufra REPUBLIC Aswan Villa Cisneros SAHARA Zouerate Tamanrasset OF EGYPT Abu Simbel Bordj Mokhtar Nouadhibou Atar Tessalit In Guezzam 20° N Port Sudan 20° N CAPE VERDE Nouakchott MAURITANIA MALI Pampa Dirkou Karima (Merowe) Tombouctou Sal Gao Agadès NIGER CHAD ERITREA Dakar SENEGAL Khartoum Kassala Asmara Tambacounda Kayès Abeché THE GAMBIA Ziguinchor Banjul Bamako BURKINA F. Niamey Ouagadougou Sokoto Zinder Geneina El Fasher El Obeid DJIBOUTI Bissau Kano Maiduguri N'Djamena Lalibela Bobo-Dioulasso BENIN SUDAN Djibouti GUINEA-BISSAU Conakry GUINEA Kankan Tamale Niamtougou Kaduna Gwasero Abuja Jos Maroua Malakal Dire Dawa Hargeisa 10° N Freetown CÔTE Bata Bida Garoua Moundou Addis Ababa SIERRA LEONE Bouaké GHANA Lagos NIGERIA N'Gaoundéré CENTRAL Gambella D'IVOIRE Accra Lomé Cotonou Calabar Mamfé Kumasi ETHIOPIA Monrovia (Roberts) Abidjan Port Harcourt Foumban AFRICAN REPUBLIC SOMALIA M'Banga Berbérati Juba Makale LIBERIA TOGO Malabo Douala Yaoundé CAM. Bangui Mandera Lodwar EQUATORIAL GUINEA KENYA Bunia UGANDA Mogadishu São Tomé Libreville CONGO Kisangani Kisimayu Makoua Entebbe 0° SÃO TOMÉ AND PRÍNCIPE Port-Gentil GABON Franceville DEMOCRATIC RWANDA Goma Nakuru Nairobi Garissa 0° Kigali Brazzaville REPUBLIC Bujumbura Mwanza Kilimanjaro Kindu Mombasa Pointe-Noire Kinshasa OF CONGO BURUNDI Dodoma Mahé Praslin Kananga Kalemie Zanzibar Mbeya TANZANIA Dar-es-Salaam SEYCHELLES Luanda Saurimo 10° S Lubumbashi ZAMBIA MALAWI COMOROS 10° S Luena Moroni (Hahaia) Kuito Bie Mfuwe Antsiranana Ndola Lichinga Mayotte Nosy-Be ANGOLA Mongu Solwezi Kapiri Lilongwe (Fr.) Cuito Cuanavale Lusaka Nampula Chileka Mahajanga EN-ROUTE RADIO NAVIGATION AIDS Livingstone Gokwe Harare Tete Ankazobe Sainte-Marie AIDES DE RADIONAVIGATION EN ROUTE Victoria Falls Hwange ZIMBABWE Quelimane Maintirano Antananarivo Toamasina Moramanga MAURITIUS 20° S AYUDAS PARA LA RADIONAVEGACIÓN EN RUTA NAMIBIA Maun Bulawayo Masvingo Beira Morondava Plaisance Francistown Saint-Denis Windhoek MOZAMBIQUE MADAGASCAR Réunion NON-DIRECTIONAL RADIO BEACON Walvis Bay BOTSWANA Greefswald Pietersburg Toliara (Fr.) Kang RADIOPHARE NON DIRECTIONNEL Gaborone Limpopo Hartebeespoortdam Tolagnaro RADIOFARO NO DIRECCIONAL Maputo Keetmanshoop Johannesburg Manzini SWAZILAND VHF OMNI-DIRECTIONAL RADIO RANGE Upington RADIOPHARE OMNIDIRECTIONNEL VHF Bloemfontein Maseru 30° S Durban -30° S RADIOFARO OMNIDIRECCIONAL VHF SOUTH DISTANCE MEASURING EQUIPMENT AFRICA LESOTHO DISPOSITIF DE MESURE DE DISTANCE Cape Town EQUIPO TELEMETRICO LOCATOR RADIOBALISE 0 275 550 825 1100 1375 1650 1925 2200 Km 40° S RADIOBALIZA DE POSICIÓN -40° S 38343 20° W 10° W 0° 10° E 20° E 30° E 40° E 50° E JANUARY 2011 Appendix 4e Total Capacity Supplied 470 Annual seats, domestic (number) Annual seats, intercontinental (number) Annual seats, total (number) Country 2001 2004 2007 2001 2004 2007 2001 2004 2007 Angola 115,960 888,160 1,199,016 304,365 354,016 588,978 817,999 1,634,414 2,272,173 Benin -- -- -- 170,653 103,233 99,268 572,433 440,222 422,400 Botswana 193,648 212,784 241,696 -- -- -- 569,634 664,794 677,404 Burkina Faso 28,362 30,622 20,245 113,707 113,698 147,095 383,656 372,975 412,061 Burundi -- -- -- -- -- 6,864 142,983 160,011 270,227 Cameroon 640,220 641,177 105,742 470,461 467,071 398,034 1,784,023 1,752,045 975,865 Cape Verde 994,179 750,889 575,120 418,349 526,617 538,094 1,620,918 1,471,323 1,168,759 Central African Republic -- -- -- 26,869 10,738 23,842 144,991 74,074 44,503 Chad -- -- -- 77,175 75,166 88,608 252,926 214,829 197,682 Comoros 27,720 94,475 111,258 183,447 223,009 336,540 351,020 466,811 571,879 Congo, Dem. Rep. 477,438 381,511 327,988 112,801 163,618 193,414 1,008,707 902,805 989,619 Congo, Rep. 512,564 806,822 443,634 52,390 106,093 117,962 953,186 1,281,700 913,478 Côte d'Ivoire -- -- -- 509,492 280,523 297,891 1,730,391 1,319,873 1,148,893 Equatorial Guinea -- -- 17,446 105,664 189,746 214,708 208,628 289,020 321,585 Eritrea 31,200 94,146 222,317 272,289 268,736 365,701 388,141 451,588 Ethiopia 932,616 887,332 728,797 745,396 1,030,439 2,004,704 2,324,225 2,879,530 4,570,648 Gabon 808,184 596,761 374,400 212,732 195,832 122,720 1,492,461 1,276,834 769,912 Gambia, The -- -- -- 95,069 66,183 126,061 378,853 229,697 252,270 Ghana -- -- 144,183 673,485 837,785 832,895 1,401,638 1,419,812 1,886,897 Guinea -- -- -- 118,460 107,354 111,462 411,925 309,764 317,378 Guinea-Bissau -- -- -- 29,744 39,910 20,280 117,468 73,684 66,846 Kenya 2,166,121 2,344,949 2,093,416 1,551,097 1,957,904 2,755,352 6,059,931 6,843,545 7,993,551 Lesotho -- -- -- -- -- -- 45,240 63,336 85,592 Liberia -- -- -- 35,681 11,440 40,040 113,165 88,712 161,485 Madagascar 1,333,890 1,324,447 1,291,284 340,535 466,665 574,379 1,865,735 2,019,011 2,147,009 Malawi 188,071 174,373 167,007 20,358 14,235 3,842 559,130 547,404 605,880 Mali 76,811 18,720 -- 207,398 198,536 165,776 831,800 822,355 730,231 Mauritania 268,437 311,792 20,475 78,600 97,708 148,928 518,561 632,853 404,257 Mauritius 232,128 321,984 502,892 1,993,050 2,085,538 2,235,545 2,675,485 2,822,664 3,282,799 Mozambique 660,417 628,793 1,144,644 61,100 74,022 91,637 1,026,731 1,059,270 1,819,116 Namibia 148,980 123,409 84,162 157,820 222,300 242,736 948,086 1,224,613 1,204,710 Niger -- -- -- 38,428 33,800 41,717 215,703 179,930 170,131 Nigeria 1,059,292 1,996,592 9,304,568 1,220,923 1,492,170 2,437,702 3,479,787 4,432,305 13,116,014 Rwanda 5,434 -- -- 16,181 18,304 18,304 225,013 279,398 485,507 São Tomé and Príncipe 11,856 -- -- 14,300 50,505 16,478 61,152 68,328 45,725 Senegal 80,219 109,538 127,244 817,893 851,897 1,231,358 1,640,869 1,803,805 2,618,012 Seychelles 651,040 680,550 712,530 331,773 392,548 424,008 1,188,031 1,218,354 1,277,562 Sierra Leone -- -- -- 36,504 30,186 48,893 132,097 165,793 228,522 Somalia 91,035 35,549 126,633 28,752 4,485 50,005 198,179 134,414 350,701 South Africa 22,240,929 22,782,903 31,767,537 5,856,431 6,443,168 7,707,063 32,352,734 34,332,229 45,789,157 Sudan 507,429 471,575 610,263 906,861 1,200,747 1,571,891 1,899,138 2,167,423 3,099,328 Swaziland -- -- -- -- -- -- 63,934 100,854 141,388 Tanzania 1,155,362 1,567,908 1,871,255 442,514 494,587 585,763 2,413,840 3,257,388 3,728,170 Togo -- -- -- 135,714 106,808 76,856 461,747 331,373 296,361 Uganda 73,983 28,392 70,980 185,484 212,940 493,740 1,059,435 929,600 1,543,057 Zambia 175,539 108,511 437,658 63,180 78,624 113,217 953,082 841,685 2,010,641 Zimbabwe 479,080 393,120 237,835 200,360 157,456 182,585 2,011,848 1,412,844 1,530,406 Low-income, fragile 769,396 876,134 930,370 638,779 772,262 1,239,498 1,848,947 2,079,351 2,834,134 Low-income, nonfragile 186,388 226,164 179,572 124,965 108,893 119,996 508,682 408,422 448,427 Middle-income 4,293,495 3,599,684 4,729,884 2,828,044 3,031,996 3,962,775 6,164,366 5,906,664 7,737,390 Resource-rich 545,598 787,085 1,561,591 379,306 462,163 628,202 1,315,692 1,543,362 2,630,742 Africa 1,607,628 1,603,779 2,052,968 891,106 986,941 1,338,270 2,437,573 2,454,204 3,327,081 471 Source: AICD data. Note: -- = not available. Intercontinental flights exclude those between North Africa and Sub-Saharan Africa. 472 Africa's Transport Infrastructure Appendix 4f Costs of Airport Construction versus Rehabilitation a. Estimated basic construction costs of a new airport with a 3,000-meter runway Running Area Length Width Area Unit of measure Costs (US$) total (US$) Terminal (2 floors) 100 100 20,000 Meters squared 53,819,552 53,819,552 Overall land requirement n.s. n.s. 6.63 Kilometers squared n.i. n.i. Apron n.s. n.s. 85,760 Meters squared 18,462,259 72,281,811 Taxiway to apron 250 21 5,250 Meters squared 1,412,763 73,694,574 Runway 3,000 21 n.s. Meters 17,716,535 91,411,110 Parallel taxiway 3,000 21 63,000 Meters squared 13,562,527 104,973,637 b. Estimated costs of rehabilitating an airport with a 2,000- by 30-meter runway, extending to 3,000 meters, and adding a parallel taxiway Item Unit cost per meter (US$) Cost Rehab 2,000 meters of asphalt 5,506 11,011,788 Add 1,000-meter extension 8,000 8,000,000 Add full-length taxiway 4,593 13,779,528 Total 32,791,316 Source: Per-unit costs are from Florida Department of Transportation, http://www.dot.state.fl.us/planning/ policy/costs/Airports.pdf. Note: n.s. = not specified; n.i. = not included. The per-unit costs have been cross-checked with estimates on currently proposed airport projects in Africa. Data do not include land acquisition costs. Also missing are other significant costs, such as a control tower, an instrument landing system, fuel facilities, vehicles, a fire station, parking facilities, landside access, and so on. Airports and Air Transport 473 Appendix 4g Domestic Air Transport Markets in Africa, 2007 Annual Estimated growth Estimated seat- seat- City pairs, Net city-pair seats kilometers kilometers Airlines November change, Country (millions) (millions) (%) (number) 2007 2004­07 South Africa 15.9 14,309.96 11.8 12 36 ­8 Nigeria 4.7 2,235.54 66.8 7 19 13 Mozambique 0.6 492.62 19.7 3 28 9 Kenya 1.0 408.13 ­3.7 4 15 ­3 Tanzania 0.9 386.24 ­1.8 5 16 ­3 Madagascar 0.6 335.71 3.7 2 24 ­61 Angola 0.6 309.64 10.0 2 21 4 Sudan 0.3 256.69 12.9 3 13 ­5 Congo, Dem. Rep. 0.2 170.91 ­5.7 2 9 ­7 Mauritius 0.3 150.47 16.0 2 1 0 Ethiopia 0.4 129.87 ­6.5 1 8 ­42 Congo, Rep. 0.2 83.85 ­18.1 4 1 ­7 Zambia 0.2 65.82 57.7 2 6 0 Botswana 0.1 64.53 6.3 1 3 ­3 Cape Verde 0.3 56.01 ­7.9 1 10 ­1 Zimbabwe 0.1 48.12 ­16.4 1 5 3 Gabon 0.2 46.51 ­9.4 1 9 ­2 Somalia 0.1 45.22 54.5 4 5 2 Namibia 0 22.21 ­12.1 1 7 ­6 Malawi 0.1 20.28 ­1.1 1 3 ­3 Ghana 0.1 18.67 -- 1 4 -- Senegal 0.1 17.38 4.0 1 3 0 Cameroon 0.1 16.90 ­49.0 3 3 ­7 Seychelles 0.4 15.45 1.5 1 1 0 Uganda 0 12.71 33.6 1 4 3 Comoros 0.1 10.94 11.9 3 7 6 Eritrea 0 9.33 -- 1 -- -- Mauritania 0 3.38 ­62.0 1 -- -- Burkina Faso 0 3.38 ­12.9 1 1 0 Equatorial Guinea 0 2.09 -- 1 1 -- Source: Analysis of data provided by Seabury ADG. Note: -- = not available. Countries are listed from highest to lowest number of estimated seat-kilometers. During the year, airlines may have stopped servicing a city pair--that is, although the Republic of Congo may show four airlines for 2007, in November 2007, there were, in fact, only two. Significant are the very high growth rates in Nigeria, Mozambique, and Zambia. Although Somalia is also growing at a very high rate, the domestic market is roughly only one-tenth the size of Kenya's, for example. Countries with missing growth rates represent new data where previous services in 2001 either did not exist or were not published. 474 Africa's Transport Infrastructure Appendix 4h Market Concentration, 2007 Herfindahl Index, domestic and Herfindahl Index, international Herfindahl Index, international Country markets domestic market market Angola 0.3 0.7 0.3 Benin 0.1 -- 0.1 Botswana 0.6 1.0 0.5 Burkina Faso 0.2 1.0 0.2 Burundi 0.3 -- 0.3 Cameroon 0.1 0.4 0.1 Cape Verde 0.6 1.0 0.4 Central African Republic 0.5 -- 0.5 Chad 0.4 -- 0.4 Comoros 0.4 0.5 0.5 Congo, Dem. Rep. 0.2 0.6 0.2 Congo, Rep. 0.3 0.9 0.2 Côte d'Ivoire 0.1 -- 0.1 Equatorial Guinea 0.1 1.0 0.1 Eritrea 0.2 1.0 0.3 Ethiopia 0.7 1.0 0.7 Gabon 0.4 1.0 0.2 Gambia, The 0.1 -- 0.1 Ghana 0.1 1.0 0.1 Guinea 0.2 -- 0.2 Guinea-Bissau 0.5 -- 0.5 Kenya 0.4 0.6 0.3 Lesotho 1.0 -- 1.0 Liberia 0.2 -- 0.2 Madagascar 0.7 1.0 0.3 Malawi 0.3 1.0 0.2 Mali 0.1 -- 0.1 Mauritania 0.2 1.0 0.2 Mauritius 0.3 0.5 0.3 Mozambique 0.3 0.5 0.2 Namibia 0.4 1.0 0.4 Niger 0.2 -- 0.2 Nigeria 0.1 0.2 0.1 Rwanda 0.3 -- 0.3 São Tomé and Príncipe 0.3 -- 0.3 Senegal 0.1 1.0 0.1 Seychelles 0.7 1.0 0.4 Sierra Leone 0.2 -- 0.2 (continued) Airports and Air Transport 475 Herfindahl Index, domestic and Herfindahl Index, international Herfindahl Index, international Country markets domestic market market Somalia 0.3 0.4 0.3 South Africa 0.2 0.2 0.1 Sudan 0.2 0.7 0.1 Swaziland 0.7 -- 0.7 Tanzania 0.2 0.3 0.1 Togo 0.2 -- 0.2 Uganda 0.2 1.0 0.2 Zambia 0.2 0.7 0.1 Zimbabwe 0.3 1.0 0.2 Low-income, fragile 0.3 0.8 0.2 Low-income, nonfragile 0.3 0.7 0.3 Middle-income 0.5 0.8 0.4 Resource-rich 0.2 0.7 0.2 Africa 0.3 0.8 0.3 Source: AICD data. Note: -- = not available. Appendix 4i Trends in Aircraft Age (seat-kilometers as percentage of total) Aircraft of unknown age Old aircraft Recent aircraft (% of total) (% of total) (% of total) Country 2001 2004 2007 2001 2004 2007 2001 2004 2007 Angola 18.6 57.8 40.2 14.1 0.2 0.1 67.3 41.9 59.7 Benin 13.1 3.6 4.2 32.0 13.7 7.4 54.9 82.7 88.5 Botswana 0.9 0.4 0.0 11.6 0.0 0.0 87.6 99.6 100.0 Burkina Faso 2.3 3.6 0.0 16.6 23.3 6.6 81.2 73.1 93.4 Burundi 37.7 13.8 0.0 46.8 65.7 12.7 15.5 20.5 87.3 Cameroon 13.2 10.2 0.3 16.4 1.5 7.9 70.4 88.3 91.8 Cape Verde 0.0 2.7 0.9 0.6 1.8 0.0 99.4 95.5 99.1 Central African Republic 15.1 72.5 0.0 27.9 0.0 0.0 57.0 27.5 100.0 Chad 0.0 13.8 0.0 9.9 0.0 0.5 90.1 86.2 99.5 Comoros 14.4 8.3 0.6 31.1 46.6 19.7 54.5 45.1 79.7 Congo, Dem. Rep. 10.2 6.5 0.6 50.1 48.6 24.7 39.7 44.9 74.7 Congo, Rep. 20.4 34.4 5.0 28.5 10.1 21.7 51.1 55.5 73.3 (continued) 476 Africa's Transport Infrastructure Aircraft of unknown age Old aircraft Recent aircraft (% of total) (% of total) (% of total) Country 2001 2004 2007 2001 2004 2007 2001 2004 2007 Côte d'Ivoire 0.6 4.0 0.5 9.0 14.6 8.7 90.4 81.4 90.8 Equatorial Guinea 3.9 1.7 1.7 1.4 0.1 0.0 94.7 98.1 98.3 Eritrea 12.2 7.7 1.5 22.1 0.0 10.7 65.7 92.3 87.8 Ethiopia 5.6 1.2 0.5 3.2 2.7 0.9 91.2 96.1 98.5 Gabon 41.1 33.3 0.8 6.7 4.1 1.2 52.3 62.6 98.1 Gambia, The 0.9 3.0 0.0 71.6 84.0 5.5 27.5 13.0 94.5 Ghana 0.7 3.9 0.6 47.3 39.7 2.6 52.0 56.4 96.8 Guinea 0.8 7.4 0.0 12.1 12.6 4.9 87.1 80.0 95.1 Guinea-Bissau 18.1 0.0 0.0 11.8 0.0 0.0 70.1 100.0 100.0 Kenya 6.5 11.2 14.1 9.8 8.8 5.7 83.7 79.9 80.2 Lesotho 0.0 0.0 0.0 0.0 0.0 0.0 100.0 100.0 100.0 Liberia 0.0 19.7 0.0 52.4 33.2 67.1 47.6 47.0 32.9 Madagascar 15.4 1.8 0.4 14.6 23.8 21.6 70.0 74.4 78.0 Malawi 5.4 1.9 0.0 56.2 64.8 20.6 38.3 33.3 79.4 Mali 2.8 9.1 0.7 10.1 9.3 3.7 87.1 81.5 95.6 Mauritania 0.0 1.2 0.0 33.7 24.2 10.5 66.3 74.7 89.5 Mauritius 18.3 0.9 5.5 0.8 0.8 0.7 80.9 98.2 93.8 Mozambique 32.9 25.8 26.7 3.9 8.1 16.3 63.2 66.0 57.0 Namibia 27.2 56.2 19.9 22.0 3.7 1.1 50.8 40.0 79.0 Niger 0.0 0.0 0.0 18.1 4.0 5.7 81.9 96.0 94.3 Nigeria 9.9 12.8 8.6 18.9 3.3 20.0 71.2 83.9 71.4 Rwanda 6.5 7.2 0.0 45.1 13.7 4.2 48.4 79.1 95.8 São Tomé and Príncipe 14.7 6.6 16.4 3.0 0.0 0.0 82.3 93.4 83.6 Senegal 12.6 0.5 0.3 3.8 5.0 1.4 83.7 94.5 98.3 Seychelles 3.0 2.1 0.0 5.7 0.4 2.9 91.3 97.4 97.1 Sierra Leone 6.1 19.6 0.0 93.9 52.2 45.1 0.0 28.2 54.9 Somalia 86.4 77.1 77.3 12.5 17.8 21.7 1.1 5.2 1.0 South Africa 18.6 13.1 14.0 18.5 5.4 2.2 62.9 81.6 83.8 Sudan 8.4 22.5 14.6 17.3 4.0 8.8 74.4 73.5 76.7 Swaziland 3.8 0.0 0.0 0.0 26.5 23.3 96.2 73.5 76.7 Tanzania 8.2 5.6 3.6 25.7 32.8 17.1 66.1 61.6 79.3 Togo 3.9 1.4 0.2 64.1 5.9 0.4 32.0 92.6 99.5 Uganda 12.9 6.3 1.2 17.3 23.3 25.7 69.9 70.4 73.2 Zambia 39.1 17.7 16.4 23.2 13.0 19.8 37.7 69.3 63.8 Zimbabwe 6.5 9.1 13.0 41.2 21.6 15.5 52.3 69.4 71.4 Low-income, fragile 7.9 5.4 3.4 21.4 18.7 9.4 70.7 76.0 87.2 Low-income, nonfragile 15.2 17.1 7.3 36.7 26.9 15.8 48.2 56.0 76.9 Middle-income 11.7 10.6 7.6 9.9 7.3 4.1 78.4 82.1 88.3 Resource-rich 17.2 22.7 9.7 15.2 4.0 8.9 67.7 73.3 81.4 Africa 12.5 12.9 6.6 21.8 15.7 9.8 65.7 71.4 83.6 Source: AICD data. Appendix 4j Trends in Aircraft Size (seat-kilometers as percentage of total) Aircraft of unknown size Large aircraft Medium-sized aircraft Smaller-sized aircraft (% of total) (% of total) (% of total) (% of total) Country 2001 2004 2007 2001 2004 2007 2001 2004 2007 2001 2004 2007 Angola 1.0 1.6 0.1 91.3 81.8 86.0 7.7 16.6 13.9 0 0 0 Benin 2.1 0 0 53.4 66.8 57.9 42.1 29.3 41.2 2.3 3.9 0.9 Botswana 0 0 0 0 0 0 0 0 0 100 100 100 Burkina Faso 0 3.6 0 81.2 60.3 45.7 11.7 17.3 46.7 7.1 18.8 7.6 Burundi 0 0 0 0 0 29.2 90.9 61.8 64.5 9.1 38.2 6.3 Cameroon 0 0 0 77.0 71.8 65.7 20.6 28.0 31.9 2.4 0.2 2.4 Cape Verde 0 0 0 75.0 56.8 6.4 17.2 37.9 88.1 7.9 5.3 5.5 Central African Republic 5.1 7.1 0 55.3 27.5 76.5 37.9 65.5 23.5 1.8 0 0 Chad 0 13.3 0 81.7 55.7 6.4 17.3 31.0 93.6 1.0 0 0 Comoros 0 0 0 51.7 29.7 70.5 45.8 65.3 25.3 2.5 5.0 4.2 Congo, Dem. Rep. 1.2 4.0 0 51.3 49.8 57.5 46.8 45.9 39.3 0.7 0.3 3.2 Congo, Rep. 12.4 26.1 0 48.4 27.3 48.2 27.4 45.8 50.5 11.9 0.7 1.3 Côte d'Ivoire 0 0.3 0 83.6 60.3 47.7 15.3 31.6 46.8 1.1 7.8 5.5 Equatorial Guinea 0 0 0 27.6 39.9 19.2 72.1 59.6 80.7 0.4 0.5 0.2 Eritrea 10.7 0 1.5 57.3 77.1 63.7 32.1 22.9 34.8 0 0 0 Ethiopia 0.4 0.4 0.1 46.7 46.5 58.6 42.8 47.3 39.7 10.0 5.7 1.5 Gabon 0 7.4 0 68.4 65.4 62.1 29.9 24.6 27.6 1.7 2.6 10.3 Gambia, The 0 0 0 70.6 74.2 9.2 10.5 15.2 88.2 18.8 10.5 2.5 Ghana 0.2 0.1 0 95.0 93.6 84.3 4.3 6.1 14.4 0.5 0.1 1.3 Guinea 0.8 4.8 0 72.7 44.8 61.4 25.1 45.8 37.8 1.4 4.5 0.8 Guinea-Bissau 0 0 0 62.3 46.6 78.7 22.3 44.7 0.0 15.4 8.7 21.3 477 (continued) 478 Aircraft of unknown size Large aircraft Medium-sized aircraft Smaller-sized aircraft (% of total) (% of total) (% of total) (% of total) Country 2001 2004 2007 2001 2004 2007 2001 2004 2007 2001 2004 2007 Kenya 0 1.5 0.1 78.1 73.1 76.7 19.1 22.2 20.8 2.8 3.2 2.4 Lesotho 0 0 0 0 0 0 0 0 0 100 100 100 Liberia 0 5.9 0 88.8 47.0 8.4 11.2 47.1 90.0 0 0 1.5 Madagascar 0 0 0 68.2 70.9 69.9 25.6 21.8 20.7 6.1 7.3 9.4 Malawi 0 0 0 12.6 5.6 21.1 76.2 84.3 71.1 11.2 10.1 7.8 Mali 1.0 8.5 0 71.9 47.0 45.9 22.7 42.0 51.8 4.4 2.5 2.3 Mauritania 0 1.2 0 52.9 17.5 0 16.6 72.3 98.1 30.5 9.0 1.9 Mauritius 0 0.9 0 95.5 89.0 92.1 2.6 8.0 6.3 1.9 2.1 1.6 Mozambique 0 0 0 52.3 55.8 43.3 37.1 33.1 42.5 10.6 11.1 14.2 Namibia 0 0 0 56.3 64.8 64.1 31.6 26.0 28.3 12.0 9.1 7.6 Niger 0 0 0 75.6 52.1 49.4 23.7 43.0 44.9 0.7 4.9 5.7 Nigeria 0.2 3 0 90.6 84.1 70.4 9.2 12.5 27.6 0.0 0.4 2.0 Rwanda 0 0 0 30.4 25.3 26.0 54.0 73.6 72.0 15.6 1.1 2.0 São Tomé and Príncipe 0 0 0 80.4 27.3 80.7 14.7 72.7 16.4 4.9 0 2.9 Senegal 1.1 0 0 85.2 60.1 60.7 11.0 36.9 38.1 2.7 3.0 1.2 Seychelles 0.9 0 0 89.5 80.4 87.4 8.1 18.5 11.5 1.5 1.1 1.1 Sierra Leone 0 0 0 0 45.6 30.5 95.6 49.3 67.1 4.4 5.1 2.4 Somalia 82.4 77.1 77.3 0 0 0 12.5 0 21.7 5.1 22.9 1.0 South Africa 0 0.3 0 70.7 69.1 67.1 26.4 26.9 29.0 2.9 3.6 3.8 Sudan 3.1 4.5 1.0 72.3 47.2 61.2 23.5 46.8 36.7 1.1 1.4 1.1 Swaziland 0 0 0 0 0 0 0 0 0 100 100 100 Tanzania 0 0.5 0 53.7 40.1 51.4 36.6 47.7 37.4 9.7 11.8 11.2 Togo 0 0 0 75.6 83.9 58.9 22.1 11.8 40.7 2.3 4.4 0.4 Uganda 3.7 1.0 0 58.7 52.9 70.9 35.1 45.0 27.9 2.5 1.1 1.2 Zambia 0 0 0.2 18.7 51.5 37.0 66.2 39.2 50.6 15.1 9.3 12.2 Zimbabwe 0 0 5.4 62.5 58.0 47.9 34.7 39.7 42.7 2.8 2.2 4.0 Low-income, fragile 0.6 1.2 0 58.5 49.4 48.7 33.5 43.6 46.8 7.4 5.9 4.5 Low-income, nonfragile 6.7 6.6 5.6 54.1 44.8 48.1 34.5 41.3 42.6 4.7 7.3 3.7 Middle-income 2.9 1.2 1.1 44.2 39.6 35.6 25.6 33.6 38.1 27.4 25.6 25.2 Resource-rich 1.8 6.2 0.2 64.0 58.3 50.7 30.4 33.8 45.9 3.7 1.7 3.3 Africa 3.1 3.6 1.9 54.7 47.2 45.6 31.3 38.8 43.3 10.9 10.4 9.1 Source: AICD data. 479 480 Africa's Transport Infrastructure Appendix 4k Safety Assessments, 2007 Country has FAA IASA safety Known carriers based registered carriers audit status in country having on EU blacklist (1 = passed, passed IATA IOSA Country (1 = yes, 0 = no) 2 = failed) audit (% of total) Angola 1 0 0 Benin 0 0 0 Botswana 0 0 0 Burkina Faso 0 0 0 Burundi 0 0 0 Cameroon 0 0 0 Cape Verde 0 1 50 Central African Republic 0 0 0 Chad 0 0 0 Comoros 1 0 0 Congo, Dem. Rep. 1 2 0 Congo, Rep. 0 0 0 Côte d'Ivoire 0 2 0 Equatorial Guinea 1 0 0 Eritrea 0 0 0 Ethiopia 0 1 100 Gabon 0 0 0 Gambia, The 0 2 0 Ghana 0 2 0 Guinea 0 0 0 Guinea-Bissau 0 0 0 Kenya 0 0 11.1 Lesotho 0 0 0 Liberia 0 0 0 Madagascar 0 0 100 Malawi 0 0 0 Mali 0 0 0 Mauritania 0 0 0 Mauritius 0 0 100 Mozambique 0 0 100 Namibia 0 0 100 Niger 0 0 0 Nigeria 0 0 28.6 Rwanda 1 0 0 São Tomé and Príncipe 0 0 0 Senegal 0 0 50 Seychelles 0 0 100 Sierra Leone 1 0 0 (continued) Airports and Air Transport 481 Country has FAA IASA safety Known carriers based registered carriers audit status in country having on EU blacklist (1 = passed, passed IATA IOSA Country (1 = yes, 0 = no) 2 = failed) audit (% of total) Somalia 0 0 0 South Africa 0 1 53.8 Sudan 1 0 20 Swaziland 1 2 0 Tanzania 0 0 33.3 Togo 0 0 0 Uganda 0 0 0 Zambia 0 0 0 Zimbabwe 0 2 33.3 Low-income, fragile n.a. n.a. 28.8 Low-income, nonfragile n.a. n.a. 2.2 Middle-income n.a. n.a. 46.2 Resource-rich n.a. n.a. 5.4 Africa n.a. n.a. 21.6 Source: AICD data. Note: EU = European Union; FAA = Federal Aviation Administration; IASA = International Aviation Safety Assessment; IATA = International Air Transport Association; IOSA = IATA Operational Safety Audit. n.a = not applicable. APPENDIX 5 Ports and Shipping 483 484 Appendix 5a Annual Traffic Container General cargo Vessel Total Vessel Exports Imports calls Exports Imports handled calls (noncontainerized (noncontainerized Port (number) (TEU) (TEU) (TEU) (number) tons) tons) Abidjan, Côte d'Ivoire -- -- -- 500,119 -- -- -- Apapa, Nigeria -- 26,348 184,047 336,308 -- 1,712,793 1,666,070 Beira, Mozambique 50 -- -- 50,000 -- 486,032 345,729 Bissau, Guinea-Bissau -- -- -- 50,000 -- -- -- Calabar, Nigeria -- -- -- -- -- 173 48,058 Cape Town, South Africa 1,037 190,904 153,924 690,895 389 336,103 223,499 Conarky, Guinea -- -- -- 753,827 -- -- -- Cotonou, Benin 410 -- -- 158,201 155 113,910 1,032,268 Dakar, Senegal 856 39,719 109,362 331,191 675 1,895,000 4,214,000 Dar es Salaam, Tanzania 478 107,107 91,365 198,472 289 645,223 1,328,529 Djibouti, Djibouti -- -- -- 195,000 -- -- -- Douala, Cameroon -- -- -- 190,700 -- -- -- Durban, South Africa 1,174 469,694 619,942 1,899,065 849 3,660,167 3,562,000 East London, South Africa 56 15,954 21,166 49,338 10 104,239 295,978 Harcourt, Nigeria -- 110 5,961 7,900 303 71,465 810,536 Lomé, Togo -- -- -- 460,000 -- -- -- Luanda, Angola 272 180,407 196,700 377,208 425 -- -- Luderitz, Namibia -- -- -- 6,154 -- -- -- Malabo, Equatorial Guinea -- -- -- -- 80 -- -- Maputo, Mozambique -- -- -- 44,000 -- -- -- Mindelo, Cape Verde -- 10,000 10,000 150,000 380 -- -- Mombasa, Kenya 491 94,120 193,223 436,671 302 139,000 1,009,000 Monrovia, Liberia -- -- -- 50,000 200 -- -- Onne, Nigeria -- 2,475 60,922 86,290 -- 97,833 1,531,983 Owendo, Gabon -- -- -- 15,942 -- -- -- Pointe Noire, Congo, Rep. -- 102,800 102,800 -- -- -- -- Port Elizabeth, South Africa 549 53,930 167,282 369,759 86 481,111 367,807 Port of Banjul, Gambia, The -- -- -- 100,000 -- -- -- Rades, Tunisia 1,230 -- -- 380,000 -- -- -- Richards Bay, South Africa 4 3,489 1,020 5,179 230 4,203,367 145,427 Saldanha, South Africa -- -- -- -- 87 1,219,448 -- Suakin, Sudan -- -- -- -- -- 50,000 102,000 Sudan, Sudan 406 170,858 25,536 328,690 409 7,686,009 1,584,181 Takoradi, Ghana -- -- -- 51,000 -- -- -- Tema, Ghana -- -- -- 420,000 -- -- -- Toamasina, Madagascar 308 12,404 26,837 92,529 168 235,067 2,071,800 Walvis Bay, Namibia 145 3,220 21,312 71,456 -- 855,007 270,373 Source: Africa Infrastructure Country Diagnostic (AICD) Database. Note: -- = not available, TEU = twenty-foot equivalent unit. 485 486 Africa's Transport Infrastructure Appendix 5b Institutional Characteristics Landlord Port has model is Management potential used (0 = no Concessions contracts as cargo landlord present in present in transshipment model; terminal infrastructure hub (0 = no; 1 = landlord operations (0 = no, Port 1 = yes) model) (0 = no, 1 = yes) 1 = yes) Abidjan, Côte d'Ivoire 1 1 1 0 Apapa, Nigeria 1 1 1 0 Bata, Equatorial Guinea 0 0 0 0 Beira, Mozambique 0 1 1 0 Boma, Congo, Dem. Rep. 0 0 0 0 Buchanan, Liberia 0 0 0 0 Calabar, Nigeria 0 -- -- -- Cap Lopez, Gabon 0 0 0 0 Cape Town, South Africa 1 0 0 0 Conarky, Guinea 1 0 1 0 Cotonou, Benin 1 0 0 0 Dakar, Senegal 1 0 0 0 Dar es Salaam, Tanzania 1 1 1 0 Douala, Cameroon 1 1 1 0 Durban, South Africa 1 0 0 0 East London, South Africa 0 0 0 0 Gamba Terminal, Gabon 0 0 0 0 Gentil, Gabon 0 1 1 0 Kamsar, Guinea 1 1 1 0 Kpeme, Togo 0 0 0 0 Kribi, Cameroon 0 0 0 0 Lomé, Togo 1 1 1 0 Luanda, Angola 1 0 0 0 Luba, Equatorial Guinea 0 0 0 0 Luderitz, Namibia 0 0 0 0 Majajanga, Madagascar 0 0 0 0 Malabo, Equatorial Guinea 0 0 -- 0 Maputo, Mozambique 0 1 1 0 Matadi, Congo, Dem. Rep. 0 0 0 0 Mindelo, Cape Verde 1 0 0 0 Mombasa, Kenya 1 0 1 0 Monrovia, Liberia 0 1 0 0 Onne, Nigeria 1 1 1 0 (continued) Ports and Shipping 487 Landlord Port has model is Management potential used (0 = no Concessions contracts as cargo landlord present in present in transshipment model; terminal infrastructure hub (0 = no; 1 = landlord operations (0 = no, Port 1 = yes) model) (0 = no, 1 = yes) 1 = yes) Owendo, Gabon 0 1 1 0 Pepel, Sierra Leone 0 0 0 0 Pointe Noire, Congo, Rep. 1 0 0 0 Port Elizabeth, South Africa 0 0 0 0 Port of Banjul, Gambia, The 0 0 0 0 Port St. Louis, Mauritius 0 -- -- -- Rades, Tunisia 0 0 0 0 Richards Bay, South Africa 0 1 1 0 Saldanha, South Africa 1 -- -- -- Sherbro, Sierra Leone 0 1 0 0 Suakin, Sudan 1 0 0 0 Sudan, Sudan 1 0 0 0 Takoradi, Ghana 1 1 1 0 Tema, Ghana 1 1 1 0 Toamasina, Madagascar 1 1 1 0 Walvis Bay, Namibia 1 1 0 0 Source: AICD Database. Note: -- = not available. 488 Appendix 5c Infrastructure Facilities Container Total length Average length Total length Water depth handling Container of general of general Mobile harbor Total berths of container of container capacity gantry cranes cargo cargo vessels cranes Port (number) berths (meters) berths (meters) (TEU per year) (number) berths (meters) (meters) (number) Abidjan, Côte d'Ivoire 26 -- 34.5 600,000 4 -- -- 2 Apapa, Nigeria 29 1,000 10.5 500,000 3 2,307 114 25 Bata, Equatorial Guinea 9 -- -- -- -- 682 -- 1 Beira, Mozambique 9 645 11 100,000 2 -- 93 2 Bissau, Guinea-Bissau 5 -- -- 100,000 -- 400 -- -- Boma, Congo, Dem. Rep. 3 -- -- -- 0 480 92 3 Buchanan, Liberia 7 -- -- -- -- 334 -- 0 Calabar, Nigeria 10 -- -- -- -- 860 -- 1 Cap Lopez, Gabon 1 -- -- 50,000 -- -- -- -- Cape Town, South Africa 34 2,234 10.7 950,000 5 2,706 138 -- Conarky, Guinea 13 269 10.5 -- -- 830 -- 50 Cotonou, Benin 11 220 10 400,000 0 990 115 1 Dakar, Senegal 52 1,273 10 400,000 0 1,245 120 3 Dar es Salaam, Tanzania 11 550 12.2 400,000 3 1,464 123 -- Djibouti, Djibouti 18 400 11 350,000 4 1,032 -- -- Douala, Cameroon 18 660 8.5 270,000 2 150 -- 0 Durban, South Africa 57 2,128 12.8 1,450,000 9 200 120 -- East London, South Africa 13 253 10.7 90,000 0 1,303 120 -- Freeport, Liberia 7 819 9.76 -- -- -- -- -- Gamba Terminal, Gabon 1 -- -- -- -- -- -- -- Gentil, Gabon 3 -- 10 100,000 -- 375 -- -- Harcourt, Nigeria 8 -- -- -- -- 1,107 -- -- Kamsar, Guinea 3 -- -- -- -- 116 -- -- Kpeme, Togo 2 -- -- -- -- -- -- -- Kribi, Cameroon 1 -- -- -- 0 250 105 -- Lomé, Togo 8 250 11.5 250,000 -- 367 -- 2 Luanda, Angola 11 448 10.5 400,000 0 825 120 -- Luba, Equatorial Guinea 1 -- -- -- -- 290 -- -- Luderitz, Namibia 8 -- -- -- -- 339 -- -- Majajanga, Madagascar 5 -- -- -- -- 700 -- 0 Malabo, Equatorial Guinea 2 -- -- -- -- -- -- 1 Maputo, Mozambique -- 300 10.3 100,000 2 1,200 135 -- Matadi, Congo, Dem. Rep. 10 520 -- 200,000 0 772 98 3 Mindelo, Cape Verde 8 315 11.5 50,000 0 700 120 0 Mombasa, Kenya 29 964 10.36 600,000 6 950 108 7 Monrovia, Liberia 8 600 9.15 100,000 0 600 -- 0 Onne, Nigeria 6 250 13.5 100,000 -- 1,590 -- -- Owendo, Gabon 3 500 9.8 100,000 0 500 -- -- Pepel, Sierra Leone -- -- -- -- -- -- -- -- Pointe Noire, Congo, Rep. -- 250 7.6 150,000 0 -- 88.5 -- Port Elizabeth, South Africa 15 635 12.2 500,000 4 1,900 144 0 Port of Banjul, Gambia, The 6 300 13 150,000 -- 120 -- 3 Port St. Louis, Mauritius -- -- -- -- -- -- -- 0 Rades, Tunisia 11 530 7.5 400,000 -- -- -- -- (continued) 489 490 Container Total length Average length Total length Water depth handling Container of general of general Mobile harbor Total berths of container of container capacity gantry cranes cargo cargo vessels cranes Port (number) berths (meters) berths (meters) (TEU per year) (number) berths (meters) (meters) (number) Richards Bay, South Africa 21 -- -- -- 0 1,244 130 2 Saldanha, South Africa 7 -- -- -- -- -- 136 0 Sherbro, Sierra Leone 0 -- -- -- -- -- -- -- Suakin, Sudan 6 -- -- -- -- 392 -- 1 Sudan, Sudan -- 420 12.6 400,000 3 2,011 115 16 Takoradi, Ghana 13 -- -- -- -- 714 105 0 Tema, Ghana 14 383 11.5 375,000 3 2,196 92 30 Toamasina, Madagascar 6 307 11 500,000 -- 526 98 3 Walvis Bay, Namibia 9 -- 10.6 100,000 0 1,426 104 1 Low-income, fragile 98 2,758 15 1,400,000 4 4,019 95 63 Low-income, nonfragile 201 6,495 11 3,890,000 22 10,850 127 3 Resource-rich 109 3,528 10 2,070,000 8 11,339 109 45 Middle-income 150 4,642 11 2,875,000 16 9,985 110 46 Africa 558 17,423 11 10,235,000 50 36,193 114 157 Source: AICD Database. Note: Country type numbers are either totals or simple averages, as appropriate. -- = not available, TEU = twenty-foot equivalent unit. Ports and Shipping 491 Appendix 5d Cargo-Handling Performance Indicators Container Container crane, vessel berth, General cargo productivity productivity vessel berth gross average average productivity (containers per (containers average (tonnes Port hour)a per hour)b per hour)b Abidjan, Côte d'Ivoire 18 35 -- Apapa, Nigeria 12 28 28 Beira, Mozambique 10 20 -- Boma, Congo, Dem. Rep. -- 6 10 Cape Town, South Africa 18 36 -- Cotonou, Benin -- -- 60 Dakar, Senegal -- 10 -- Dar es Salaam, Tanzania 20 19.8 22.6 Djibouti, Djibouti 17 68 -- Douala, Cameroon 18.5 37.5 12 Durban, South Africa 15 45 -- East London, South Africa 8 10 -- Freeport, Liberia -- 8 -- Harcourt, Nigeria -- -- 26 Kribi, Cameroon -- -- 5.5 Lomé, Togo -- 14 -- Luanda, Angola 6.5 7 16 Luderitz, Namibia -- 7.5 -- Majajanga, Madagascar -- -- 18 Malabo, Equatorial Guinea -- -- 11 Maputo, Mozambique 11 22 22.5 Matadi, Congo, Dem. Rep. 6.5 10 11 Mindelo, Cape Verde 6.5 13 13.5 Mombasa, Kenya 10 10 20.8 Onne, Nigeria -- 14 26 Pointe Noire, Congo, Rep. 6.5 6.5 13.5 Port Elizabeth, South Africa 14.95 15 45 Port of Banjul, Gambia, The -- 1.5 -- Rades, Tunisia -- 10 -- Richards Bay, South Africa -- 10 25 Suakin, Sudan -- -- 15 Sudan, Sudan 8 20 24.5 Takoradi, Ghana -- -- 30 Tema, Ghana 13 39 40 Toamasina, Madagascar -- 17.6 27 Walvis Bay, Namibia -- 7.5 31 (continued) 492 Africa's Transport Infrastructure Container Container crane, vessel berth, General cargo productivity productivity vessel berth gross average average productivity (containers per (containers average (tonnes Port hour)a per hour)b per hour)b Low-income, fragile 12.3 12.4 10.5 Low-income, nonfragile 13.2 22.2 28.6 Resource-rich 10.3 18.8 16.5 Middle-income 12.8 19.8 30.1 Africa 12.2 18.9 22.1 Source: AICD Database. Note: Country type numbers are reported as simple averages. -- = data not available. (a) Containers loaded and unloaded per single crane working hour. (b) Containers loaded and unloaded alongside berth. Appendix 5e Port Access and Landside Quality Container Truck processing Working Container vessel Cargo vessel Cargo time for receipt Adequate rail access dwell time, pre-berth Container vessel pre-berth vessel stay, and delivery of road access (0 = no, average waiting time, stay, average waiting time, average cargo, average Port (0 = no, 1 = yes) 1 = yes) (days) average (days) (hours) average (hours) (hours) (hours) Abidjan, Côte d'Ivoire -- 1 12 1 1 2.9 2.2 2.5 Apapa, Nigeria 0 1 42 12 24 36 40.8 6 Bata, Equatorial Guinea -- -- -- -- -- 6 -- -- Beira, Mozambique 1 1 20 7 24 8 48 6.8 Bissau, Guinea-Bissau 1 1 21 2.5 -- -- -- -- Boma, Congo, Dem. Rep. 0 0 -- -- -- 108 84 -- Buchanan, Liberia 1 1 -- -- -- -- -- -- Calabar, Nigeria 0 0 -- -- -- 26.4 55.2 -- Cap Lopez, Gabon -- -- 12 2 1 2 -- -- Cape Town, South Africa 0 1 6 3 24 3 36 4.8 Conarky, Guinea 0 1 15 2.5 2 -- 2.7 -- Cotonou, Benin 0 1 12 24 36 48 48 6 Dakar, Senegal 0 1 7 18 24 24 60 5 Dar es Salaam, Tanzania 1 0 7 24 24 6 62.4 5 Djibouti, Djibouti 1 -- 8 1 1 1.5 -- 12 Douala, Cameroon -- -- 12 1.6 3.2 -- -- 12 Durban, South Africa 0 1 4 5 32 2 60 5 East London, South Africa 1 1 7 0 24 6 48 2 (continued) 493 494 Container Truck processing Working Container vessel Cargo vessel Cargo time for receipt Adequate rail access dwell time, pre-berth Container vessel pre-berth vessel stay, and delivery of road access (0 = no, average waiting time, stay, average waiting time, average cargo, average Port (0 = no, 1 = yes) 1 = yes) (days) average (days) (hours) average (hours) (hours) (hours) Freeport, Liberia -- -- 15 2 1 2.5 3 5 Gamba Terminal, Gabon -- -- -- -- -- -- -- -- Gentil, Gabon 1 -- 10 1.5 1 2 2.5 -- Harcourt, Nigeria 0 1 -- -- -- 38.4 45.6 -- Kamsar, Guinea 1 1 -- -- -- -- -- -- Kpeme, Togo -- 1 -- -- -- -- -- -- Kribi, Cameroon 1 0 -- -- -- 24 72 -- Lomé, Togo 0 1 13 1 1 -- -- 4 Luanda, Angola 0 1 12 96 48 144 60 14 Luba, Equatorial Guinea 1 -- -- -- -- -- -- -- Luderitz, Namibia 1 -- 8 0 2 0 -- 3 Majajanga, Madagascar 0 -- -- -- -- 24 67.2 -- Malabo, Equatorial Guinea 1 -- -- -- -- 7 2 -- Maputo, Mozambique 1 1 22 3 24 12 55.2 4 Matadi, Congo, Dem. Rep. 0 0 25 -- 48 108 84 18 Mindelo, Cape Verde 1 0 16 24 48 36 72 6 Mombasa, Kenya 0 1 5 12 36 36 48 4.5 Monrovia, Liberia 1 -- 15 1.5 1 2.5 -- 5.5 Onne, Nigeria 0 0 30 4 32 6 38.4 24 Owendo, Gabon -- 1 10 1.5 1 2 -- -- Pepel, Sierra Leone -- -- -- -- -- -- -- -- Pointe Noire, Congo, Rep. 0 1 18 38.4 48 43.2 60 12 Port Elizabeth, South Africa 1 1 6 3 24 5 48 4.5 Port of Banjul, Gambia, The 0 0 2 2 1.5 3 -- 2.5 Port St. Louis, Mauritius 0 0 -- -- -- -- -- -- Rades, Tunisia 1 1 10 2.5 2 1.5 2 6 Richards Bay, South Africa 1 1 -- 0 24 2 36 0 Saldanha, South Africa 1 1 -- -- -- 2 28.8 -- Sherbro, Sierra Leone -- -- -- -- -- -- -- -- Suakin, Sudan 0 1 -- -- -- 28.8 57.6 -- Sudan, Sudan 0 1 28 2.1 45.6 28.8 52.8 24 Takoradi, Ghana 0 1 -- -- -- 3 52.8 -- Tema, Ghana 0 1 25 12.35 32 9.6 48 8 Toamasina, Madagascar 0 1 8 12 12.25 24 52.8 3.5 Walvis Bay, Namibia 1 1 8 0 48 0 57.6 3 Low-income, fragile -- -- 14.8 1.8 7.9 37.8 35.2 6.3 Low-income, nonfragile -- -- 8.1 3.9 22.9 5.4 43.2 4.6 Resource-rich -- -- 19.3 17.7 22.6 28.2 44.3 15.3 Middle-income -- -- 13.3 14.0 26.5 19.5 54.2 5.4 Africa -- -- 13.9 9.5 20.6 21.3 45.5 7.3 Source: AICD Database. Note: Country type figures are reported as simple averages. Vessel stay is equivalent to turnaround time. -- = data not available. 495 496 Africa's Transport Infrastructure Appendix 5f Average Port Costs and Charges (US$/unit) General cargo Bulk dry Bulk liquid Container handling handling handling cargo handling charge charge charge Port charge per TEU per ton per ton per ton Abidjan, Côte d'Ivoire 260 13.5 5 -- Apapa, Nigeria 155 8 -- 1 Bata, Equatorial Guinea 185 17 8 -- Beira, Mozambique 125 6.5 2.5 0.75 Boma, Congo, Dem. Rep. -- 10 -- 4 Cap Lopez, Gabon 280 14 -- -- Cape Town, South Africa 258.2 -- 6.3 0.4 Cotonou, Benin 180 8.5 5 -- Dakar, Senegal 160 15 5 4 Dar es Salaam, Tanzania 275 13.5 4.5 3.5 Djibouti, Djibouti 135 7.5 4 1.25 Douala, Cameroon 220 6.5 6 -- Durban, South Africa 258.2 8.4 1.4 -- East London, South Africa -- 8.4 6.3 -- Freeport, Liberia -- 8 5.5 3.3 Gentil, Gabon 280 11 -- -- Harcourt, Nigeria -- 8 -- 1 Kribi, Cameroon -- 12 -- -- Lomé, Togo 220 9 5 -- Luanda, Angola 320 8.5 5 -- Luderitz, Namibia 90 12 5 3 Majajanga, Madagascar -- 6 -- -- Malabo, Equatorial Guinea 185 17 -- -- Maputo, Mozambique 155 6 2 0.5 Matadi, Congo, Dem. Rep. 120 10 8 2 Mindelo, Cape Verde 100 10 5 2.5 Mombasa, Kenya 67.5 6.5 5 -- Monrovia, Liberia 200 10.5 4 -- Onne, Nigeria 145 6.5 -- -- Owendo, Gabon 340 16 -- -- Pointe Noire, Congo, Rep. 140 5.5 2.75 -- Port Elizabeth, South Africa 258.2 8.4 1.4 -- Port of Banjul, Gambia, The 210 13 5 -- Rades, Tunisia -- 9 4.5 -- Richards Bay, South Africa -- -- 1.4 -- (continued) Ports and Shipping 497 General cargo Bulk dry Bulk liquid Container handling handling handling cargo handling charge charge charge Port charge per TEU per ton per ton per ton Suakin, Sudan -- -- 3 1 Sudan, Sudan 150 10 3 1 Takoradi, Ghana -- 7 2.5 1.5 Tema, Ghana 168 10 3 1.5 Toamasina, Madagascar -- 6 3 -- Walvis Bay, Namibia 110 15 5 2 Low-income, fragile 210 10.1 4.9 3.3 Low-income, nonfragile 172.8 9.8 4 1.8 Resource-rich 226 11.2 5 1 Middle-income 161.5 8.5 3.6 2 Africa 191.7 9.9 4.3 1.9 Source: AICD Database. Note: Country type figures are reported as simple averages. Charges are ship to gate or rail. -- = data not available, TEU = twenty-foot equivalent unit. APPENDIX 6 Expenditure Needs Appendix 6 Costs of Achieving Targets of Pragmatic Scenario, by Expenditure Purpose (US$, millions/year) CAPEX Improve Upgrade Expand Total Total Country OPEX condition category capacity CAPEX spending Angola 140 39 57 5 101 241 Benin 33 15 23 0 39 72 Botswana 47 12 12 0 24 72 Burkina Faso 60 3 41 0 45 105 Burundi 17 9 4 0 13 30 Cameroon 97 19 78 2 99 196 Cape Verde 3 3 0 3 6 9 Central African Republic 69 8 82 0 91 160 Chad 102 3 79 53 135 237 Congo, Rep. 55 22 21 8 52 106 Congo, Dem, Rep. 352 45 350 129 524 876 Côte d'lvoire 130 27 56 2 84 215 Equatorial Guinea 19 8 3 17 27 46 (continued) 499 500 Africa's Transport Infrastructure CAPEX Improve Upgrade Expand Total Total Country OPEX condition category capacity CAPEX spending Eritrea 13 7 2 0 10 23 Ethiopia 127 12 73 21 106 233 Gabon 52 23 36 1 60 112 Gambia, The 15 17 3 1 20 36 Ghana 98 43 40 3 87 184 Guinea 77 19 55 3 77 154 Guinea-Bissau 13 8 2 0 10 23 Kenya 130 66 40 11 117 247 Lesotho 11 3 2 0 5 16 Liberia 38 8 45 8 61 99 Madagascar 154 18 73 8 99 252 Malawi 34 9 16 0 26 60 Mali 83 16 36 18 70 153 Mauritania 39 16 8 16 40 80 Mauritius 14 8 13 6 27 40 Mozambique 167 21 78 5 104 271 Namibia 67 39 18 0 57 124 Niger 67 9 10 6 24 91 Nigeria 373 151 67 61 279 652 Rwanda 11 3 3 3 9 20 Senegal 49 20 15 3 38 87 Sierra Leone 32 9 17 6 32 64 South Africa 421 182 60 46 288 710 Sudan 221 17 146 148 311 532 Swaziland 29 4 10 3 16 46 Tanzania 223 26 135 16 178 401 Togo 18 12 10 25 47 65 Uganda 55 14 57 3 74 130 Zambia 134 21 32 0 53 186 Zimbabwe 69 23 16 1 40 110 Resource-rich 1,192 302 520 294 1,116 2,307 Low-income, nonfragile 1,329 293 648 114 1,056 2,385 Low-income, fragile 844 192 642 174 1,009 1,853 Middle-income 593 251 115 59 425 1,018 Africa 3,958 1,039 1,925 641 3,605 7,563 Source: Carruthers, Krishnamani, and Murray 2009 Note: OPEX = operating expenditure; CAPEX = capital expenditure. APPENDIX 7 Financing 501 502 Appendix 7a Transport Spending and Finance Sources, by Country (annualized flows) GDP share (%) US$ millions/year O&M Capital expenditure O&M Capital expenditure public Public Non-OECD Total Total public Public Non-OECD Total Total Country sector sector ODA financiers PPI CAPEX spending sector sector ODA financiers PPI CAPEX spending Benin 0.97 0.37 1.07 0.14 0.00 1.58 2.55 41 16 46 6 0 68 109 Botswana 1.01 1.33 0.00 0.09 0.00 1.42 2.43 106 140 0 9 0 149 256 Cameroon 1.06 0.19 0.32 0.08 0.00 0.59 1.64 175 32 53 13 0 98 273 Cape Verde 6.04 2.30 4.45 0.14 0.00 6.89 12.92 61 23 45 1 0 69 130 Chad 0.08 0.23 0.39 0.09 0.00 0.72 0.79 4 14 23 5 0 42 46 Congo, Dem. Rep. 0.43 -- 0.77 0.02 0.00 -- 0.43 30 -- 55 2 0 -- 30 Congo, Rep. 1.47 2.29 0.28 0.20 0.00 2.77 4.23 89 139 17 12 0 168 258 Côte d'lvoire 0.04 0.80 0.02 0.06 0.08 0.97 1.00 6 131 3 10 14 158 164 Ethiopia 0.43 1.74 1.70 0.16 0.00 3.61 4.04 53 215 209 20 0 444 497 Ghana 1.19 0.47 1.11 0.08 0.01 1.66 2.85 127 50 119 8 1 178 305 Kenya 1.50 0.42 0.77 0.07 0.12 1.38 2.88 280 79 144 13 22 259 539 Lesotho 0.80 0.59 1.29 0.32 0.00 2.21 3.01 11 8 18 5 0 31 43 Madagascar 2.41 0.64 2.60 0.08 0.16 3.48 5.89 122 32 131 4 8 175 297 Malawi 1.74 0.87 1.01 0.08 0.00 1.97 3.72 50 25 29 2 0 56 106 Mali 0.18 0.80 1.49 0.45 0.09 2.82 3.00 10 42 79 24 5 150 159 Mozambique 1.06 0.73 1.61 0.25 0.85 3.44 4.50 70 48 106 16 56 226 296 Namibia 0.98 0.22 0.37 0.08 0.00 0.67 1.65 61 14 23 5 0 42 103 Niger 0.23 0.47 0.89 0.28 0.00 1.63 1.86 8 16 30 9 0 54 62 Nigeria 0.10 0.69 0.02 0.48 0.38 1.57 1.67 112 780 26 537 422 1,765 1,877 Rwanda 0.27 0.14 1.06 0.14 0.00 1.34 1.61 6 3 25 3 0 32 38 Senegal 0.08 0.86 0.54 0.26 0.05 1.71 1.79 7 75 47 23 5 149 156 South Africa 1.93 0.76 0.00 0.00 0.18 0.94 2.87 4,661 1,843 0 0 444 2,287 6,948 Tanzania 1.37 0.58 0.82 0.05 0.06 1.52 2.89 194 82 117 7 9 214 408 Uganda 0.51 0.07 0.86 0.02 0.26 1.21 1.71 44 6 75 2 22 106 150 Zambia 1.35 1.79 0.93 0.11 0.04 2.86 4.21 99 132 68 8 3 210 309 Middle-income 1.88 0.78 0 03 0.01 0.16 0.98 2.86 5,081 2,103 88 22 444 2,657 7,738 Resource-rich 0.32 0.74 0.11 0.34 0.21 1.39 1.72 720 1,646 234 745 469 3,095 3,815 Low-income, nonfragile 0.98 0.67 1.12 0.22 0.12 2.13 3.11 1,084 737 1,241 242 128 2,347 3,431 Low-income, fragile 0.16 0.56 0.61 0.13 0.04 1.33 1.49 60 214 234 49 14 511 571 Africa 1.20 0.74 0.28 0.16 0.16 1.34 2.54 7,701 4,724 1,797 1,059 1,055 8,635 16,336 503 504 Appendix 7b Potential Efficiency Gains GDP share(%) US$ millions/year Capital Tariff cost Capital Tariff cost Country Undercollection execution recovery Total Undercollection execution recovery Total Benin 0.00 0.00 0.01 0.01 0 0 0 0 Botswana -- 0.26 -- 0.26 -- 27 -- 27 Cameroon 0.20 0.06 0.00 0.26 33 9 0 43 Cape Verde -- 0.30 -- 0.30 -- 3 -- 3 Chad 0.14 0.04 0.25 0.43 8 2 15 25 Congo, Dem. Rep. -- 0.00 -- 0.00 -- 0 -- 0 Congo, Rep. 0.00 0.54 0.08 0.62 0 33 5 38 Côte d'lvoire -- 0.23 0.20 0.42 -- 37 32 59 Ethiopia 0.22 0.38 0.30 0.91 28 47 37 112 Ghana 0.32 0.00 0.39 0.71 34 0 42 76 Kenya 0.00 0.10 0.14 0.24 0 19 25 44 Lesotho 0.08 0.00 0.67 0.75 1 0 10 11 Madagascar 0.00 0.19 0.52 0.71 0 10 26 36 Malawi -- 0.31 0.79 1.10 -- 9 22 31 Mali -- 0.00 -- 0.00 -- 0 -- 0 Mozambique 0.47 0.21 0.19 0.87 31 14 13 57 Namibia 0.56 0.13 0.46 1.16 35 8 29 72 Niger 0.17 0.07 0.59 0.83 6 2 20 28 Nigeria -- 0.20 -- 0.20 -- 223 -- 223 Rwanda 0.00 0.04 0.16 0.20 0 1 4 5 Senegal -- 0.25 -- 0.25 -- 21 -- 21 South Africa -- 0.23 -- 0.23 -- 553 -- 553 Tanzania 0.34 0.19 0.00 0.53 48 27 0 75 Uganda -- 0.02 -- 0.02 -- 2 -- 2 Zambia 0.00 0.48 0.27 0.75 0 35 20 55 Middle-income 0.01 0.23 0.01 0.25 36 613 38 688 Resource-rich 0.08 0.20 0.07 0.35 170 455 163 788 Low-income, nonfragile 0.15 0.15 0.19 0.49 169 162 211 541 Low-income, fragile -- 0.16 0.12 0.28 -- 61 46 107 Africa 0.08 0.20 0.09 0.37 497 1,298 574 2,368 Source: Derived from Foster and Briceño-Garmendia 2009. Note: -- = not available. 505 506 Africa's Transport Infrastructure Appendix 7c Agency and Total Social Benefits of Timely Road Maintenance The financial analysis in chapter 8 attempted to identify the extent to which shortfalls in the financing of maintenance and investment in trans- port infrastructure, and roads in particular, might be overcome by more efficient policies. It has been widely recognized for many years that inad- equate funding of current and periodic maintenance might both increase total road agency costs (because of the extra burden of rehabilitation and reconstruction expenditure that it caused) and substantially reduce road user welfare (by increasing vehicle operating costs). Most traditional research on the topic has concentrated on the total social benefit (the sum of user benefits and agency benefit), while the balance between these two types of benefits has been given less attention. The importance of that bal- ance in the present context is that any agency benefits from changed poli- cies accrue directly to the public budget, and the potential saving to road users is available to finance the policy change only if part or all of the ben- efits can be captured for the public purse by changes in road user charges. To examine this balance more closely, we compared the total life-cycle costs of two possible maintenance and rehabilitation road agency alterna- tives for a typical road class in Africa. The road agency alternatives that were evaluated are to (i) perform annual routine maintenance and reha- bilitate the road when its condition becomes poor and (ii) perform annual routine maintenance and proper periodic maintenance. The evaluation was done using the Highway Development and Management Model (HDM-4) using representative vehicle fleet characteristics for Africa. Sixteen road classes were evaluated, corresponding to two climate/ terrain types, four traffic levels, and two road condition types. The two climate/terrain types evaluated were dry and flat roads and wet and hilly roads. The four traffic levels were 500, 1,000, 3,000 or 6,000 annual aver- age daily traffic (AADT). Roads with 500 and 1,000 AADT were consid- ered surface treatment roads, and roads with 3,000 and 6,000 AADT were considered asphalt concrete roads. Each road was designated as being in either good or fair condition. For the surface treatment roads, the following road agency alternatives were evaluated: · Rehabilitation: Perform annual routine maintenance and rehabilitate the road to a surface treatment standard when the road roughness reaches International Roughness Index (IRI) value of 10 m/km1 Financing 507 · Periodic maintenance: Perform annual routine maintenance and periodic maintenance corresponding to 12 millimeter (mm) reseals done every seven years. If during the evaluation period the road roughness reaches 10 IRI, m/km, the road is rehabilitated to a surface treatment standard. For the asphalt concrete roads, the following road agency alternatives were evaluated: · Rehabilitation: Perform annual routine maintenance and rehabilitate the road to an asphalt concrete standard when the road roughness reaches 10 IRI, m/km. · Periodic maintenance: Perform annual routine maintenance and peri- odic maintenance corresponding to 50 mm overlays done when the road reaches 4 IRI, m/km. Table A7c.1 presents the unit cost of road works adopted on the evalua- tion that corresponds to representative figures for Africa. The performance of each road class under the two possible alternatives was evaluated with HDM-4 for a 20-year evaluation period, and the resulting road agency costs and total transport costs (road agency costs plus road user costs) were calculated, both undiscounted and discounted using a 12 percent discount rate. Table A7c.2 shows that the ratio between the undiscounted costs of total rehabilitation and the undis- counted costs of periodic maintenance varies from 0.42 to 2.92. The table also shows that when all costs are discounted to a present value at 12 percent a year, the ratio between the rehabilitation alternative costs and the periodic maintenance alternative costs varies from 0.49 to 1.85. Table A7c.2 considered only the road life-cycle agency costs in financial terms. Table A7c.3 presents the present value, at a 12 percent Table A7c.1 Unit Cost of Road Works Surface type Type of road work Unit cost (US$/km) Surface treatment Routine maintenance 2,600 Periodic maintenance (reseal 12 mm) 24,150 Reconstruction surface treatment pavement 273,000 Asphalt or concrete Routine maintenance 2,600 Periodic maintenance (overlay 50 mm) 73,500 Reconstruction asphalt or concrete pavement 315,000 Source: World Bank Rocks Database. 508 Table A7c.2 Comparison of Road Agency Costs for Various Roads and for Rehabilitation vs. Periodic Maintenance Undiscounted Present value Undiscounted rehabilitation/ Present value rehabilitation/ value of road agency periodic agency costs periodic Road class Alternative costs (US$/km) maintenance ratio (US$/km) maintenance Dry/flat 500 AADT Rehabilitation 325,000 2.61 53,448 0.87 ST road fair condition Periodic maintenance 124,450 61,767 Dry/flat 1,000 AADT Rehabilitation 325,000 2.61 91,823 1.49 ST road fair condition Periodic maintenance 124,450 61,767 Dry/flat 3,000 AADT Rehabilitation 367,000 1.84 102,604 0.94 AC road fair condition Periodic maintenance 199,000 108,679 Dry/flat 6,000 AADT Rehabilitation 367,000 1.84 148,974 1.39 AC road fair condition Periodic maintenance 199,000 107,240 Wet/hilly 500 AADT Rehabilitation 325,000 0.82 100,232 0.99 ST road fair condition Periodic maintenance 397,450 101,528 Wet/hilly 1,000 AADT Rehabilitation 325,000 0.87 132,011 1.17 ST road fair condition Periodic maintenance 373,300 112,686 Wet/filly 3,000 AADT Rehabilitation 367,000 1.84 135,343 1.21 AC road fair condition Periodic maintenance 199,000 112,095 Wet/filly 6,000 AADT Rehabilitation 367,000 1.84 164,241 1.47 AC road fair condition Periodic maintenance 199,000 112,095 Dry/flat 500 AADT Rehabilitation 52,000 0.42 21,751 0.49 ST road good condition Periodic maintenance 124,450 44,457 Dry/flat 1,000 AADT ST road Rehabilitation 325,000 2.61 53,448 1.20 good condition Periodic maintenance 124,450 44,457 Dry/flat 3,000 AADT AC road Rehabilitation 367,000 2.92 67,629 1.58 good condition Periodic maintenance 125,500 42,881 Dry/flat 6,000 AADT AC road Rehabilitation 367,000 2.92 79,300 1.85 good condition Periodic maintenance 125,500 42,881 Wet/hilly 500 AADT ST road Rehabilitation 52,000 0.42 21,751 0.49 good condition Periodic maintenance 124,450 44,457 Wet/hilly 1,000 AADT ST road Rehabilitation 325,000 2.61 66,283 1.49 good condition Periodic maintenance 124,450 44,457 Wet/hilly 3,000 AADT AC road Rehabilitation 367,000 2.92 67,629 1.58 good condition Periodic maintenance 125,500 42,881 Wet/hilly 6,000 AADT AC road Rehabilitation 367,000 1.84 102,604 1.77 good condition Periodic maintenance 199,000 57,814 Source: Calculations by R Archondo Callao using HDM4. Note: ST = surface treatment; AC = asphalt concrete. 509 510 Table A7c.3 Total Economic Costs Discounted at 12 Percent Total value of Rehabilitation/ agency costs periodic Road class Alternative (US$/km) maintenance ratio NPV (US$/km) NPV/PV of agency costs Dry/flat 500 AADT ST Rehabilitation 658,946 1.04 24,586 0.40 road fair condition Periodic maintenance 634,360 Dry/flat 1,000 AADT ST Rehabilitation 1,851,305 1.04 69,264 1.12 road fair condition Periodic maintenance 1,782,041 Dry/flat 3,000 AADT AC Rehabilitation 3,589,946 1.08 277,785 2.56 road fair condition Periodic maintenance 3,312,161 Dry/flat 6,000 AADT AC Rehabilitation 6,952,731 1.07 436,891 4.07 road fair condition Periodic maintenance 6,515,840 Wet/hilly 500 AADT ST Rehabilitation 761,213 1.00 3,771 0.04 road fair condition Periodic maintenance 757,442 Wet/hilly 1,000 AADT ST Rehabilitation 2,080,735 1.01 18,177 0.16 road fair condition Periodic maintenance 2,062,558 Wet/hilly 3,000 AADT AC Rehabilitation 4,027,216 1.07 264,184 2.36 road fair condition Periodic maintenance 3,763,031 Wet/hilly 6,000 AADT AC Rehabilitation 7,937,865 1.07 502,055 4.48 road fair condition Periodic maintenance 7,435,810 Dry/flat 500 AADT ST Rehabilitation 561,097 0.97 ­19,450 ­0.44 road good condition Periodic maintenance 580,547 Dry/flat 1,000 AADT ST Rehabilitation 1,695,824 1.02 41,015 0.92 road good condition Periodic maintenance 1,654,809 Dry/flat 3,000 AADT AC Rehabilitation 3,341,756 1.04 114,690 2.67 road good condition Periodic maintenance 3,227,066 Periodic maintenance 6,405,968 Wet/hilly 500 AADT ST Rehabilitation 646,004 0.98 ­15,981 ­0.36 road good condition Periodic maintenance 661,984 Wet/hilly 1,000 AADT ST Rehabilitation 1,935,770 1.02 34,134 0.77 road good condition Periodic maintenance 1,901,636 Wet/hilly 3,000 AADT AC Rehabilitation 3,776,335 1.03 111,081 2.59 road good condition Periodic maintenance 3,665,253 Wet/hilly 6,000 AADT AC Rehabilitation 7,481,304 1.02 145,671 2.52 road good condition Periodic maintenance 7,335,633 Source: Calculations by R Archondo Callao using HDM4. Note: ST = surface treatment; AC = asphalt concrete. 511 512 Africa's Transport Infrastructure discount rate, of the total economic costs (sum of road agency costs and road user costs) as well as the net present value (NPV) of the periodic maintenance alternatives and the ratio between the NPV and the present value of financial agency costs. The NPV is the dif- ference between the present value (PV) of total society costs of the base alternative (rehabilitation) and the project alternative (periodic maintenance). The optimal alternative from an economic point of view is the one with the lowest present value of total society costs. The periodic mainte- nance alternative is the recommended alternative if the rehabilitation/ periodic maintenance ratio is higher than 1.0. The study shows that only for the two roads with traffic of 500 AADT and in good condition is the recommended alternative to perform routine maintenance and rehabili- tation, but no rehabilitation is needed during the evaluation period in both cases. For all other cases, proper periodic maintenance is the recom- mended alternative. These results can be interpreted as follows. First, performing timely periodic maintenance is likely to save on agency costs if no discounting is applied, except where traffic volumes are as low as 500 vehicles per day. When discounting is applied at 12 percent, timely routine maintenance may or may not reduce total agency costs, depending on the combination of climate/terrain, traffic volumes, and initial surface condition. When user costs are taken into account, however, the only circumstance in which there is a net loss from timely maintenance is when traffic volume is low and the initial condition is good; clearly this circumstance will change as surface conditions deteriorate from good to fair. Moreover, only a small proportion of roads meet those criteria. While the cases included in this examination represent only a small set of the very wide range of possible situations, it can be safely concluded that timely periodic main- tenance, whether or not it reduces road agency costs, will yield high net social benefits, which may be up to five times the extra expenditures on maintenance of the more heavily trafficked roads. Appendix 7d Closing the Gap (US$ millions/year) Gain from eliminating inefficiencies Gain from Total Gain from cost recovery Gain from Infrastructure spending raising capital through ending Financing gap Country spending needs in sector Total gain budget execution fuel levy undercollection or surplus Benin 116 105 0 0 0 0 (11) Botswana 107 107 27 27 -- -- 27 Cameroon 328 220 43 9 0 33 (65) Cape Verde 19 19 3 3 -- -- 3 Chad 364 46 25 2 15 8 (292) Congo, Dem. Rep. 1,474 87 0 0 -- -- (1,387) Congo, Rep. 163 163 38 33 5 0 38 Côte d'lvoire 341 164 69 37 32 -- (107) Ethiopia 393 298 112 47 37 28 17 Ghana 307 301 76 0 42 34 71 Kenya 474 474 44 19 25 0 (44) Lesotho 25 25 11 0 10 1 11 Madagascar 377 297 36 10 26 0 44 Malawi 85 85 31 9 22 -- 31 Mali 236 154 0 0 -- -- (83) Mozambique 395 296 57 14 13 31 (41) Namibia 154 103 72 8 29 35 21 (continued) 513 514 Gain from eliminating inefficiencies Gain from Total Gain from cost recovery Gain from Infrastructure spending raising capital through ending Financing gap Country spending needs in sector Total gain budget execution fuel levy undercollection or surplus Niger 139 62 28 2 20 6 (50) Nigeria 1,222 842 223 223 -- -- (157) Rwanda 36 29 5 1 4 0 (2) Senegal 141 94 21 21 -- -- (25) South Africa 7,971 6,948 553 553 -- -- (470) Tanzania 619 408 75 27 0 48 (136) Uganda 221 150 2 2 -- -- (70) Zambia 280 239 55 35 20 0 14 Middle-income 8,430 7,738 688 613 38 36 (4) Resource-rich 3,810 3,113 788 455 163 170 91 Low-income, nonfragile 3,797 3,266 541 162 211 169 10 Low-income, fragile 3,155 571 107 61 46 -- (2,477) Source: derived from Foster and Briceño-Garmendia, 2009 and Carruthers, Krishnamani, and Murray 2009. Note: -- = not available. Financing 515 Note 1. The International Roughness Index, IRI, is a measurement of the accumulated deviation of an actual surface from a true planar surface. A value of 10 meters per kilometer represents a very badly damaged pavement with many shallow and some deep depressions. Index Boxes, figures, maps, notes, and tables are indicated with b, f, m, n, and t following the page number. A availability of urban public transport, 246 Abidjan bus rapid transit bus rapid transit, 237, 239, 241, 245, availability and quality of, 247 251, 256, 260 fares, 251, 253, 256, 400 vehicle regulation, 263 financing, 256, 259 walking, 238 fleets, 241, 245 Abuja Treaty (June 1991), infrastructure, 239 179n11, 368 maintenance, 248 accidents. See safety city transport authority, 236 and security driver licensing, 260 Accra taxes on minibus owners, 250 bus rapid transit vehicle regulation, 262, 263, 264 fares, 254, 258 walking, 238, 265 financing, 257, 259 ADS-B (automatic dependent fleets, 244, 245 surveillance-broadcast), 141 infrastructure, 235b, 237, 239 Aeronautical Information Manual, 145 maintenance, 248 Aéroports de Paris, 147 operating costs, 250 Aéroports du Cameroun, 147 fuel costs, 249 AFCAC (African Civil Aviation rail commuter services, 231 Commission), 169, 370 vehicle regulation, 263 AfDB. See African Development Bank walking, 238 Africa Infrastructure Country Diagnostic ADB (Asian Development Bank), 308 (AICD), vii Addis Ababa future research, 364 air traffic growth, 153, 154, 155 517 518 Index Africa Infrastructure Country Diagnostic agriculture, 11, 22, 36, 36f, 280, 305, 306f (AICD), (continued) AICD. See Africa Infrastructure Country rail study, 366 Diagnostic road study, 18, 25, 29, 48, 50b, aid. See economic assistance; official 56, 60 development assistance (ODA) urban roads, 228, 229f Air Afrigiya, 156, 158 African Airlines Association, 174 Air Afrique, 153, 155, 156, 156b, 158, 172 African and Malgache Civil Aviation Air Algérie, 156 Authorities, 175 Air France, 153, 156b, 172 African Central Bank, 368 Air Gabon, 155, 172 African Civil Aviation Commission Air Malawi, 158 (AFCAC), 169, 370 airports and air transport, 139­79 African Development Bank (AfDB), 338, airfares, 167­68, 168f 368, 404 airport charges, 146, 146f African Economic and Monetary airport conditions, 143­45 Union, 368 airport infrastructure, 139­52 African Economic Community, 368 air traffic control (ATC) and navigation, African Railways Union, 88­89 140­42 African Union, 5­6, 14, 169, 211 in connectivity model African Union of Public Transport national, 288 (UATP), 370 regional, 287, 287t African Union of Railways, 370 financing reform, 177 Agence Nationale de Nouveaux Chemins fleet composition, 162, 162t, de Fer, 231 163­64f, 165t Agence pour la Sécurité de la Navigation "freedoms of the air," 178n10 Aérienne en Afrique et à governance and, 366­67 Madagascar, 151 investment in maintenance rather than agenda for action, 381­406, 386t construction, 176 aid dependence, 396­97 liberalization and Yamoussoukro process, airports and air transport, 390­91, 392t 167, 168­74, 170­71t, 173t, 177, attitudes toward private supply and 179n11 profit, 383­84 market structure, 159­62 behavioral issues, 397­400, 401t dominant airlines, 159­60t, 159­61 corruption, 397­99 international transport within Africa, critical transport policy issues, 382­94 159­61, 160t expenditures, 400­405 monopoly in low-density markets, 161 human resource capacity, 397 national flag carriers, 152, 158, maintenance financing, 399 161­62, 172, 176­77 mobilizing competition, 382­83 passenger seats, 159, 159t modern logistics systems, 385 operations, 152­76 mode-specific policy issues, 386­94 ownership and management, 147­50, multimodal issues, 382­86 148­49t, 366­67 overcoming institutional weaknesses, recommendations for future policies, 395­97, 398t 176­77, 390­91, 392t ports and shipping, 391­93, 393t regulatory institutions, 150­51, 177 railways, 389, 390t runways regulation, 396 capacity, 142­43 roads, 386­88, 388t condition of, 143­45, 145t, 176, safety and security, 385­86 281­82, 284 urban transport, 393­94, 395t safety, 174­75, 175f, 176, 178n8 AGETIPs (contract management regional safety oversight bodies, agencies), 46 151­52 Index 519 service and connections, 139­40, 178n6 roads and corridors service quality, 163­67 budget allocation, 52 terminal capacity, 143, 144t conditions, 29 traffic rates, 152­58, 152t road funds, 42 domestic air transport, 158 rural roads, 22 intercontinental traffic, 153­54, 154m trade routes, 19 international traffic within Africa, bicycles, use of, 230, 238 154­56, 155f, 157m, 157t bilateral trade agreements, 75­76 North African market, 156­58 Bofinger, H. C., 277 Airports Company South Africa, 150 Bolloré group (French company), 115, Air Tanzania, 158, 178n9 222n4 Angola bonds economic assistance to, 339, 352n3 corporate bond market, 343­44 independence of, 2 railway use of, 106, 135n10 oil exports, 193­94 Botswana private investment, 342 expenditures on transport infrastructure, railways, 88, 89 317, 318, 329 "Angola mode" of barter, 339, 352n3 paved roads, 20 APM Terminals, 201­2, 222n4 quality of governance, 357, 385 Arndt, C., 377n2 railways Asian Development Bank (ADB), 308 costs per traffic unit, 126 Association of Southern African National freight traffic, 98, 129 Road Agencies, 370 network density, 86 ASYCUDA, 197, 198, 223n6 rolling-stock productivity, 119 automatic dependent surveillance- residual funding gap, 347 broadcast (ADS-B), 141 trade routes, 19 AviAssist, 174 Botswana Railways, 107 Brazil bus rapid transit, 235b B transport infrastructure, spending Bamako on, 49, 294 bus rapid transit, 237, 239, 245, bribery. See corruption 251, 253 British Airports Authority, 150 city transport authority, 236 British Airways, 153, 159 modes of public transit, 236 Buchanon (Liberian port), 5b Banjul Accord Group (BAG), 140, 152, Bullock, R., 277, 288 155­56, 172, 177n3, 179n12 Burkina Faso Barret, F., 246, 253 binational railway with Côte d'Ivoire, Beitbridge Railway, 112 8, 107 Belgian Congo, independence of, 2 railways, 4, 342 Belgium and colonial interests roads and corridors in Africa, 2 conditions, 29 Benin over- and underengineering, 63 air traffic, 166 road funds, 42 expenditures on transport rural roads, 22, 33 infrastructure, 317 trade routes, 19 fuel taxes, 41 Burundi oil exports, 193, 194 road density, 20 railways road funds, 45 labor productivity, 117 Business Action for Improving network structure, 87 Customs Administrations track conditions, 88 in Africa, 197 520 Index bus rapid transit systems cartelization of trucking industry, 388 availability of, 246, 246t CCFB (railway operator), 115 bus fleets, 241­45, 244t CEAR. See Central East African Railways comparison with rail fares and travel Company times, 98, 99f CEMAC. See Economic and Monetary dedicated infrastructure for, 235b, Community of Central Africa 238­39 Central Africa fares, 250­56, 268n5 air safety record, 174 affordability, 254­56, 255t air traffic, 153, 166 levels, 253­54, 253t lake transport, 186b setting and controlling, 251, 252t, 400 railways, 86 financing and subsidies, 256­60 roads and corridors, 18, 20 large vs. small, 239­41, 240t, 258b cartelization, 388 minibuses, 236­37, 239­45, 242­45t, trade facilitation, 76 247, 248, 250 transport price, 75 overloading, 247, 249 transport quality, 73­74, 74t quality of, 246­47 Central African Republic regulation, 260 air traffic, 166 taxes, 250 roads and corridors vehicle maintenance, 248­49 density, 20 wait times, 247 paved roads, 20 spending, 60 trade routes, 19 C spending to improve infrastructure, 298 CAAs. See civil aviation authorities Central East African Railways Company Cameroon (CEAR), 8, 92, 95, 113, 115, 122, airports 124, 125, 136nn20­21 charges, 146 Chad ownership and management, 147 air traffic, 166 air traffic, 166 expenditures on transport bus rapid transit, 237 infrastructure, 312 expenditures on transport exports, value of, 6 infrastructure, 317 oil exports, 193, 194 local government networks, size of, 46 residual funding gap, 347 oil exports, 193 roads and corridors private investing in, 342 investment programs, 54, 55 railways, 84, 107, 124 maintenance expenditures, 57, 58 road freight transport, 72 over- and underengineering, 63 Camrail (Cameroon), 87, 95, 115, 119, paved roads, 20 122, 125, 131, 132, 135n14 trade routes, 18­19 Canada spending to improve infrastructure, 298 railways China derailments, 125 economic assistance from, 339, 346, network density, 86 352n3, 404 Cape Verde energy interests in Africa, 193 air safety standards, 174 investment in railways, 2, 84, 89 expenditures on transport infrastructure, transport infrastructure, spending on, 312, 318, 329 296, 312 residual funding gap, 347 civil aviation authorities (CAAs), 142, capital fund disbursement, 324­25, 324t 150­51, 175, 177, 390­91 capital investment civil conflicts and wars, 2­3, 6, 85, 122 roads, 50­56, 51f, 56f Clinton, Bill, 14 Index 521 coal exporting, 191 residual funding gap, 347 colonialism and independence, 1­3 roads and corridors, 26, 29 Comair, 159 connectivity commercial environment, 365 air transport, 139­40, 140m, 154­55, Commission for Africa, 14 155f, 158, 178n6 Commission for Africa Report, 337 category of infrastructure, 281­83, 282t Commission Internationale du Bassin condition of infrastructure, 283­84 Congo-Oubangui-Sangha, 186b insights from analysis, 306­8 Common Market for Eastern and Southern model, 272­306, 273f Africa (COMESA), 152, 172, airports, 287, 287t, 288 179n12, 198, 370 application of, 284­93 community arrangements costs of extending network and for roads, 47 increasing facilities, 292­93 commuter propeller aircraft, 166 costs of maintaining infrastructure, competition 293 mobilizing, 382­83 costs of upgrading infrastructure railway concessions, 106 standards, 292 railways competing with road sector, 105 geographic information system (GIS), Competition and Fair Trading Act 276, 280 (Zambia), 113 inputs, 275­84 Conakry macroeconomic database, 275­76 bus rapid transit, 239, 251 mixing scenarios, 299b city road authority, 232 outputs, 293­306 walking, 238 physical infrastructure needed to concessions achieve targets, 284­85, 286t ports and shipping, 199­201, 203b, ports and shipping, 287­88 214­15, 215f, 222n4 railways, 288 railways. See railways roads, 276­77, 286­87, 288, 289 roads, 50b scenarios, 284, 402 Congo, Democratic Republic of unit costs, 277, 293 oil exports, 193 urban areas, 289­90, 289­90t railways ports and rail connections, problems gauge lines, 86 in, 8 labor productivity, 117 rail routes, transcontinental, 89 mineral extraction and, 89 rural market connectivity, 306, 307f rolling-stock productivity, 119 spending to improve, 271­310 tariff rates, 101 affordability, 294­96 traffic, 95 cost estimation, 290­93, 401­2 residual funding gap, 347 databases used in study, 275­77 roads and corridors spending needs by country and coun- conditions, 29 try group, 300­301, 300t, 302t paved roads, 20 spending needs by purpose of spend- rural roads, 22 ing, 301­3, 303f, 304t traffic volumes, 26 spending needs by type of connectiv- trade clearance, 198 ity, 303­6, 304f, 305t, 306f trade routes, 19 spending needs in base scenario, 294, Congo, Republic of 295t air traffic, 166 spending needs in pragmatic scenario, expenditures on transport infrastructure, 296­99, 297t, 298f 317, 318, 329 standards, 278­81 oil exports, 193 national, 279, 288 railways, 101, 135n15 regional, 278­79, 286­88, 308n4 522 Index connectivity (continued) infrastructure, 237, 239 rural, 279­80, 289, 290t regulation, 260 urban, 280­81, 289­90, 290t city transport authority, 236, 262 container trade. See ports and shipping modes of public transit, 236 Cooperative Development of Operational road conditions, 228 Safety and Continuing train service, 97, 231, 238 Airworthiness Programme Dakar Dem Dikk (DDD), 259, 262 (COSCAP), 151 Dar es Salaam Corredor de Desenvolvimento do Norte air traffic control, 141 (CDN; railway operator), 115, 125 bus rapid transit corridor systems, development of, 8. availability and quality of, 247 See also roads fares, 251, 256 corruption, 5­6, 357­59, 358b, 376­77, fleets, 241 397­99 infrastructure, 235b, 237, 239 Corruption Perceptions Index, 357, 385 capacity and demand, 206, 208 costs city transport authority, 236, 262 airfares, 167­68, 168f public-private partnership in airport fuel costs, 249, 249t management, 150 high costs, poor service, and reduced regulation, 262 trade, 6­10 road conditions, 228 ports and shipping, 212­19 transshipment, 188 cargo-handling costs, 218­19 defined-contribution pension scheme, railways, 88 353n4 cost-sharing arrangements, 47 derailments, 125 Côte d'Ivoire deregulation airports and air transport air transport, 168­74, 170­71t, charges, 146 173t, 177 traffic, 166 road administration, 47, 79 civil war, 122, 190 Djankov, S., 358 expenditures on transport Doing Business Indicators, 360 infrastructure, 312 domestic air transport rates, 158 French influence, 3 domestic public finance, 335­37 fuel taxes, 41 Donner, M. L., 212 oil exports, 193 Douala ports and shipping bus rapid transit concessions, 222n4 fares, 251, 256 customs, 197 financing, 257, 259 private investing in, 342 infrastructure, 237, 239 railways, 4, 107, 122, 137n28 regulation, 260 binational railway with motorcycles, 238 Burkina Faso, 8 population, 225 roads and corridors, 19, 32, 72 walking, 238 stock exchange raising capital, 344 Douala International Airport, 147 customs, 9, 196­98 DP World, 189, 201, 203, 214, 222n3 Drewry Shipping Consultants, 391 D driver licensing, 260 Dakar dry bulk ship traffic, 191, 192t, 218 bus rapid transit Durban fares, 251, 253, 256, 259 capacity and demand, 206, 208 financing, 257, 259 cargo-handling rates, 183, 216 fleets, 241, 244 transshipment, 189­90 Index 523 E Eritrea air traffic, 166 East Africa railways, 86 air transport Ethiopia safety, 174, 175 air safety standards, 174 traffic control, 141 air traffic control, 141 traffic growth, 155 expenditures on transport infrastructure, bus rapid transit, 241, 244, 251 329 ports and shipping fuel taxes, 40­41 capacity and demand, 206 roads and corridors constraints on economies of scale, 194 conditions, 29 container traffic, 182, 188 economic assistance, 397 customs reform and modernization, funds, 230 198 investment programs, 54 dwell times, 216 over- and underengineering, 63 general cargo-handling performance, rural roads, 22, 33 217 safety, 264­65 planning and reform, 195 traffic volumes, 26 railways, 8, 84, 106 Ethiopian Airways, 153, 155, 159, 160, roads and corridors 161, 172 trade routes, 19 ethnic divisions, 2­3, 6 transport price, 75 European Commission, 338, 404 transport quality, 74t European Union's economic assistance, 397 East African Community (EAC), 19, 172, expenditures. See costs; financing 179n12, 197, 198 Export-Import Bank, 353n3 Economic and Monetary Community of exports, value of, 6 Central Africa (CEMAC), 151­52, 172, 186b economic assistance F aid dependence, 371­73, 396­97 fares. See specific types of transport from China, 339 financing, 311­54 costs of capital, 346 accuracy of needs estimates, 351 natural resources used to pay for, 339, airports and air transport, 177 352n3 bus rapid transit systems, 256­60 non-OECD sources, 339 expenditures, 312­22 official development assistance. See offi- agenda for action, 400­405 cial development assistance (ODA) balance between finance needs and economic communities, regional, 369­70 commitments, 317­22, 319­22f Economic Community of West African balance between investment and States (ECOWAS), 179n12 maintenance, 316­17, 318f, 351 EDIFACT (Electronic Data Interchange for categories of, 314 Administration, Commerce, and improving budget execution ratios, Transport), 223n6 351 education system, 373­75, 397 roads, 47­71, 48f. See also roads efficiency improvements, 323 sources of funding, 314­16, 315t, EgyptAir, 153, 156 316­17f, 403 elites, role of, 361­62 total spending, 312, 313t, 403 Emirates (airline), 153, 159 official development assistance. See offi- Equatorial Guinea cial development assistance (ODA) oil exports, 193, 194 private investors. See private investors private investment in recommendations for future policies, seaports, 342 350­52 524 Index financing, (continued) rolling-stock productivity, 119 residual funding gap, 347­49, 348­49f traffic trends, 92 shortfalls, dealing with, 322­47 The Gambia allocation of funds to maintenance, air traffic, 166 325­27, 325f spending to improve infrastructure, 299 capital fund disbursement, 324­25, Gautrain (South Africa), 238 324t geographic information system (GIS), input collection of user charges, 323 in connectivity model, 276, 280 costs of capital and sources of finance, Ghana 345­46, 346f air domestic public finance, 335­37, 336t airport charges, 146 efficiency improvements, 323 air traffic control, 141 increasing funding, 334­35 traffic rates, 166 local sources of finance, 342­45, 344t customs, 197 lower standards, 330­31 economic development, 3 most promising ways to increase educational attainment, 373 funds, 346­47 ethnic divisions, 3 non-OECD financing, 338­40, fuel taxes, 40 346, 350 local sources of finance, 343 reallocation in transport sector, National Transport Safety Council, 362 327­30, 328f, 329t, 330f, 331t ports and shipping rehabilitation backlogs, 331­34 landlord port model, 199 for rehabilitation backlogs, 331­34 oil exports, 193 fleet composition private investment, 342 air transport, 162, 162t, 163­64f, 165t quality of governance, 357 bus rapid transit systems, 241­45, 244t railways, 117 road transport, 74­75 roads and corridors France conditions, 29, 32 colonial interests in Africa, 2 funds, 45, 230 independence of African colonies, 3 over- and underengineering, 63 official development assistance responsibility for maintenance, 232 from, 338 rural roads, 33 freight. See also railways safety, 37, 265, 388 road transport, 71­76 trade routes, 19 French Civil Aviation Authority, 174 urban population, 308n4 fuel costs, 249, 249t urban transport, responsibility for, 232 fuel taxes, 39­41, 40f, 71, 323 Ghana Airways, 153, 155, 158, 172 Ghana Gateway Program, 188 global positioning system (GPS), 141 G governance, 355­79 Gabon aid dependence, 371­73 air traffic, 166 capacity, 371­75, 377 economic assistance to, 339 corruption. See corruption fuel taxes, 41 defined, 356 oil exports, 193 elites, role of, 361­62 private investing in, 342 financial capacity railways constraints, 371­73 concessions, 107 coordination improvements, 377 freight traffic, 98 human resource capacity, 373­75, 397 gauge lines, 86 institutional reform, 376 labor productivity, 117 institutions, 363­71 network density, 86 international institutions, 367­71 Index 525 national governance, 356­57 Industry Safety Strategy Group (ISSG), national transport institutions, 363­67. 174, 175 See also specific type of transport Infrastructure Consortium for Africa, 337 Pan-African institutions, 368­69 inland waterway transport, 186b Pan-African transport associations, 370 instrumented landing system (ILS), 141, ports and shipping, 367 145, 178n5 railways, 108­9, 366­67 intercontinental air traffic rates, 153­54, recommendations for future policies, 154m 375­77, 394­400 International Civil Aviation Organization regional economic communities, 369­70 (ICAO) regional transport associations, 370­71 Cooperative Development of roads, 364­65 Operational Safety and Continuing role of state and private sector, 359­61 Airworthiness Programme safety culture, absence of, 362­63 (COSCAP), 151, 175 state capture, 361­62 enforcement powers, 370 traditions and attitudes, 357­63 safety analysis, 174­75, 175f urban transport, 367 Universal Safety Oversight Program, 174 Group of Eight Gleneagles Summit, 337 upgrading infrastructure standards based Guinea on, 145, 169, 292 independence from France, 3 International Container Terminal Services political fragmentation, political costs Inc., 201 of, 4, 5b International Development Association railways, 86 (IDA), 137n28, 338, 346 roads and corridors, 19, 29, 32 international institutions, 367­71 Guinea-Bissau International Labour Organization oil exports, 194 (ILO), 210 Gulf states. See Persian Gulf states International Maritime Organization (IMO), 210­11 H International Monetary Fund (IMF), 347, 359 higher education, 373­74 International Ship and Port Facility Highway Development and Management Security Code (ISPS Code), Model (HDM4), 283, 309n7, 326 211­12 Hine, J., 72 international transport within Africa history, 1­16 air traffic, 154­56, 155f, 159­61, 160t colonialism and independence, 1­3, 360 railways. See railways country diversity and, 10m, 10­14, 12­13t road corridors, 18­20 distorted sector as consequence of, 3­6 Internet, customs use of, 197 high costs, poor service, and reduced investment trade, 6­10 balance between investment and new-millennium renaissance, 14­15 maintenance, 316­17 of railways, 4, 83­85 private investors. See private investors human resource capacity, 373­75, 397 railways, 89­91 roads, capital investment, 50­56, I 51f, 56f ICAO. See International Civil Aviation Organization J independence, period of, 1­3 India Japan derailments, 125 official development assistance from, 338 economic assistance from, 339, 346, 404 transport infrastructure, spending on, roads, spending on, 49 294 526 Index Jeppesen (Jeppesen Sanderson, Boeing Kruk, C. B., 212 subsidiary), 145, 178n7 Kumar, A. J., 246, 253 jets, 162, 162t, 164f, 166 Johannesburg bus rapid transit system, L 238­39 Johannesburg International Airport, 142, Lagos 143, 145, 150, 154, 155 airport, 143, 166, 167m bus rapid transit air pollution, 248 K dedicated infrastructure, 235b Kampala fares, 253, 256 bus rapid transit, 239, 248, 250, 256 financing, 256, 259 motorcycles, 238 midibuses, 237, 244 road conditions, 228, 230 minibuses, 239 Kenya city transport authority, 236 air traffic control, 141 fuel costs, 249 corruption, 358b motorcycles, 238 customs, 197 population, 225 driver regulation, 264 ports and shipping, 200, 214 expenditures on transport rail commuter services, 231, 238 infrastructure, 317 railways, 97, 238 import duty on vehicles, 250 vehicle regulation, 263, 264 private investing in, 342 lake transport, 186b railways, 84, 109, 112, 125, 131 landlocked countries roads and corridors bonded warehouses, 9 road funds, 42, 45 corridors to the sea, 18, 19, safety, 37, 265 184, 185t vehicle overloading, 69 freight transport prices, 72 trade clearance, 197 roads, spending on, 49, 51t Kenya Airways, 153, 155, 160, 161, 172 land transports, coordination with ports Kenya Bus Service (KBS), 239, 258 and shipping, 8, 222, 393 Kigali Lesotho bus rapid transit, 239, 256 expenditures on transport infrastructure, modes of public transit, 236 317, 329 population, 225 fuel taxes, 40 road conditions, 228 local sources of finance, 343 Kinshasa roads and corridors availability of urban public transport, conditions, 32 246 density, 20 bus rapid transit maintenance expenditures, 57 fares, 256 over- and underengineering, 63 financing, 256 rural connectivity, 279 large-bus service, 239 Liberia minibuses, 241 spending to improve infrastructure, overloading, 247 60, 299 population, 225 liquid bulk ship traffic, 191­94, rail commuter services, 231, 238 193f, 218 road conditions, 228 Logistics Performance Index, 385 vehicle regulation, 263 low-income countries Korea, Republic of capital road projects, 50, 51t transport infrastructure, spending on, fragile countries, 10m, 11, 12­13t 49, 296 nonfragile countries, 10m, 11, 12­13t Index 527 M residual funding gap, 347 roads and corridors, 19, 72 macroeconomic database, input in management connectivity model, 275­76 of airports, 147­50, 148­49t Madagascar port management models, 199, 200t expenditures on transport of state-owned railways, 108­9 infrastructure, 312 Maputo Corridor Development, 186 fuel taxes, 40­41 maritime transport. See ports and shipping local government networks, size of, 46 Marxism, 3, 4 ports and shipping Maseru International Airport (Lesotho), cargo-handling charges, 219 145 oil exports, 194 Matthee, M., 6 private investment, 342 Mauritania railways, 115, 342 air traffic, 166 residual funding gap, 348 economic assistance to, 339 roads and corridors oil exports, 193 economic assistance, 397 railways, 86 investment, 55 road improvement, 382 maintenance expenditures, 57 spending to improve infrastructure, 298 rural roads, 36 Mauritius spending to improve ports and shipping, 342 infrastructure, 298 roads and corridors Madarail, 9, 95, 115, 124, 130, 132, conditions, 29, 33 137n28 rural road conditions, 33 Maersk Line, 194 traffic volumes, 26 Magadi Soda Works, 109 transshipment, 189­90 Malawi Meredith, M., 2 air traffic control, 141 Metro Mass Transit Ltd. (MMT), 259, expenditures on transport infrastructure, 268n9 317 Metrorail (South Africa), 231 exports, value of, 6 middle-income countries, 10m, 11, 12­13t national construction council, 374 capital road projects, 50, 52 railways road maintenance expenditures, 59 concessioned, 107 mines, railways to, 84, 87, 89, 98, 104, 109 financial performance, 125 modern logistics systems, 385 passenger traffic, 97 Mombasa private investment, 342 air traffic control, 141 regulation, 113 port and shipping residual funding gap, 347 capacity and demand, 206, 208 roads and corridors cargo-handling systems, 209 conditions, 29 container traffic, 182 rural road conditions, 33 trade port, 200 safety, 264 trade routes, 19 spending on, 49, 387 transshipment, 188 traffic volumes, 26 monopoly trade routes, 19 for government revenues, 384 Malaysia in low-density air transport markets, 161 roads, spending on, 49 railways, 114 Mali UK airports, 150 air traffic, 166 Morocco economic development, 92 air hub for international traffic within railways, 4, 8, 87, 342 Africa, 158 528 Index motorcycles, use of, 238 national flag carriers, 152, 158, 161­62, Mozambique 172, 176­77 air transport, domestic growth of, 158 national fragmentation, 4 economic development, 92 National Freight Logistics Study, 384 expenditures on transport National Ports Authority Act (South infrastructure, 315 Africa), 206 independence of, 2 National Road Safety Council (Ghana), 37 private investing in, 342 national standards of connectivity, 279, 288 railways national transport institutions, 363­67 concessions, 115, 124 National Transport Safety Council financial performance, 125 (Ghana), 362 track conditions, 88 Naudé, W., 6 traffic, 92 new-millennium renaissance, 14­15 roads and corridors New Partnership for Africa's Development, concessions, 50b 368, 369, 404, 405 over- and underengineering, 63 Niger whole port concessions, 201 air traffic, 166 multimodal issues, 382­86 fuel taxes, 40 Mundy, M., 184, 201, 277 residual funding gap, 348 roads and corridors density, 20 N maintenance expenditures, 58 Nairobi over- and underengineering, 63 airport terminal capacity, 143, 145 rural roads, 33 air traffic control, 141 trade routes, 19 air traffic growth, 153, 154, 155 Nigeria bus rapid transit air transport and airports fares, 254, 256, 258 domestic growth, 158 financing, 259 international growth within Africa, large vs. small buses, 258b 160 midibuses, 244, 245 safety, 175 minibuses, 237, 241 traffic control, 141 overloading, 247 traffic growth, 153, 155 private operation, 239 Bureau of Public Enterprises, 203 private sector, 257 economic assistance to, 339 taxes, 250 educational attainment, 373 rail commuter services, 231, 238 ethnic divisions, 3 road conditions, 230 expenditures on transport infrastructure, Namibia 312, 315, 318, 335 corporate bond market, 343 local government networks, size of, 46 local sources of finance, 343 ports and shipping ports and shipping, planning and cargo-handling charges, 219 reform, 195 concessions, 199, 203b, 214, 340 quality of governance, 357 customs, 197 railway traffic trends, 92 landlord port model, 199 roads and corridors oil exports, 193, 194 conditions, 29 reform, 203 density, 20 railways, 84, 107, 117 road funds, 42 roads and corridors rural roads, 22, 33 concessions, 50b spending on, 387 density, 20 national classified roads, 20­22 maintenance expenditures, 58 Index 529 over- and underengineering, 63 Pan-African institutions, 368­69 traffic volumes, 26 Pan-African rail network, 88­89 stock exchange raising capital, 344 Pan-African Road Safety Conference, 37, trade clearance, 198 387 urban population, 308n4 Pan-African transport associations, 370 Nigeria Airways, 155, 172 passenger traffic Nkrumah, Kwame, 3, 360 air transport, 159, 159t NLPI (railway operator), 115 buses. See bus rapid transit systems North Africa railways, 95­97 air transport profitability, 127­28, 128f, 137n26 airport charges, 146 urban transport, 246­47, 246t traffic growth, 155, 156­58 pedestrians. See walking in cities Nyerere, Julius, 3, 355, 360 Penfold, A., 184, 201, 277 Persian Gulf states O air traffic with Africa, 153­54 economic assistance from, 339, 346, 404 official development assistance (ODA) Petit Train Bleu (Dakar), 97, 231 dependence on, 321 Pinard, M. I., 44 levels of, 335, 350 piracy, 211 non-fragile vs. fragile states, 315 political instability, 375 not fungible, 317 ports and shipping, 181­224 reallocating, 404 changing trade patterns, 181­95 roads and, 315 constraints on growth, 184­88, 187t, subsidy included, 346 189f sustaining, 337­38, 351­52, 404 container trade flows, 188 oil exports, 181­82 container traffic, 182­83f, 182­84 liquid bulk traffic, 191­94, 193f container transshipment, 188­90, Oman, C., 377n2 191f Organisation Commune Benin-Niger des dry bulk traffic, 191, 192t Chemins de Fer et des Transports economies of scale, 194­95 (OCBN), 87, 126 general cargo traffic, 190 Organisation for Economic Co-operation liquid bulk traffic, 191­94, 193f and Development (OECD), 312, concessions, 199­201, 203b, 214­15, 404 215f, 222n4 Organization of African Unity, 89, 368 in connectivity model Ouagadougou national, 288 availability of urban public transport, regional, 287­88 246 coordination with land transports, 8, bus rapid transit, 239, 251 222, 393 modes of public transit, 236 infrastructure development, 206­12, overloading 282, 284 bus rapid transit systems, 247, 249 capacity and demand, 206­8, 207t trucks, 69­70, 365 cargo-handling systems, 208­10, 213f, ownership and management. See also 223n7 concessions; private investors port development, 208, 209t of airports, 147­50, 148­49t safety and security arrangements, state ownership, 4­5. See also national 210­12 flag carriers inland waterway transport, 186b institutional and regulatory framework, P 195­206, 367 Pan-African Infrastructure Development customs, 196­98 Fund, 345 landlord ports, 199 530 Index ports and shipping, (continued) comparison with bus fares and travel planning and reform, 195­96, 221­22 times, 98, 99f port management models, 199, 200t compensation for financially unviable private sector participation, 200­203, passenger service, 133 201f, 202t concessions regulatory arrangements, 203­6, competitive nature of, 106 204­6t financing, 112, 129­32, 130­31f, service ports, 199 136n22, 137n31, 389 whole port concession, 199 future role of, 132­33, 389 performance, cost, and quality, 212­19 history of, 84, 107, 108m, 135n11 cargo-handling costs, 218­19, 220f institutional arrangements, 109­15, dry and liquid bulk cargo-handling 110­11t performance, 218 operational performance, 122­24, general cargo-handling performance, 123f 216­17, 217f, 218t shareholdings, 114­15, 116t landside container-handling termination of, 114 performance, 215­16, 216f transparency of, 359 overall quality of port services, 219 in connectivity model quayside container-handling national, 288 performance, 212­15, 215f regional, 288 rail connections, 8, 98­99, 222 consistency recommendations for future policies, across modes in policy toward 219­22, 391­93, 393t infrastructure use, 134 regional management associations, 371 in government policies, 134 strategic planning for ports, 221 decline of, 84 Portugal and colonial interests in Africa, 2 financial performance, 125­32 Precision Air, 166 concession financing, 129­32, private investors, 340­42 130­31f attitudes toward, 383 costs of repairs, 88 in bus rapid transit, 257 costs per traffic unit, 126, 126f, making infrastructure more attractive 136n24 for, 352 disaggregated real profitability ports and shipping, 200­203, 201f analysis, 126­27 in railway concessions, 115 freight service profitability, 128­29, privatization, 150, 383 129f public-private partnerships (PPPs) passenger service profitability, airports, 147, 148t 127­28, 128f, 137n26 container traffic, 184 government inspection, 135n13 roads, 45 history of, 4, 83­85 institutional arrangements, 106­15, 366 Q concessions, 109­15 regulatory framework for concessions, Qatar Airways, 159 112­14 state-owned railways. See state- R owned, this heading Raballand, G., 74 investment and maintenance, 89­91 radar, air traffic control's use of, 141, financing difficulties on low-volume 177­78n4 lines, 90 RAI. See Rural Accessibility Index investment needs, 90­91 railways, 83­138 links between, 8 binational, 8 market sector performance, 91­106 China's investment in, 2, 84, 89 freight tariffs, 101­4, 102­3f, 135n7 Index 531 freight traffic, 98­100, 100f, railways 135nn8­9 concessions, 112­14 general characteristics, 92 private rail operators, 134 passenger fares, 97­98, 400 urban transport, 260­65, 261t, 267 passenger traffic, 95­97 rehabilitation requirements traffic trends, 92­95, 93­94f, 96f, finance, 331­34 134n3 railways, 88 uncompetitive, 104­6, 105f roads, 54­55, 54f, 55t network structure, 85m, 85­89, 283 Reli Assets Holding Company (Tanzania), condition of systems, 87­88 113 density of, 86 residual funding gap, 347­49 expansion proposals, 88­89 resource-rich countries, 10m, 11, 12­13t size of, 86 Richards Bay, South Africa, 190, 191, 196, traffic volume, 87 216, 366 operational performance, 115­25 Rio Tinto Zinc, 5b concessioned railway companies, RITES company (India), 107, 114, 115, 122­24 136n17 labor productivity, 117, 118f Rizet, C., 72 rolling-stock productivity, 117­22, Road Network Evaluation Tool (RONET), 120­21f, 137n25 32, 53, 58, 59, 62, 79n1, 327, 364, service and safety, 124­25 387 port-rail connections, 8, 98­99, 222 roads, 17­81 private operators' capacity community arrangements for, 47 and willingness to provide track concessions and toll roads, 50b renewal, 133 in connectivity model, 276­77 recommendations for future policies, national, 288 132­34, 389, 390t regional, 286­87 reform, 107 rural, 289 regulation of private rail operators, 134 foreign-based firms in road construction state-owned, 105, 108­9, 113, 125 industry, 374 urban transport, 97, 231­32, 238, 268n3 freight transport services, 71­76, 74t Railway Systems of Zambia, 125 governance and, 364­65 Rastogi, C., 279 institutional arrangements, 71, 78, regional economic communities (RECs), 364­65 368­70 network, 18­29 regional roads in international context, density, 20­26, 21f, 23­24f 22­26 international transit corridors, regional standards of connectivity, 278­79, 18­20, 19t 286­88, 308n4 region's roads in international regional transport associations, 370­71 context, 22­26 regulation rural transport infrastructure, 22 agenda for action, 396 as share of GDP, 25, 25f airport infrastructure, 150­51, 177 traffic volumes, 26­29, 27­28f constraints of, 9 paved roads, 20, 25, 26t national transport institutions, 363­67 performance and conditions, 29­37, 77, ports and shipping, 195­206 281, 283, 283t customs, 196­98 commercially structured road planning and reform, 195­96, authorities and, 78 221­22 GDP per capita and, 67 port management models, 199, 200t geographic conditions and, 69, 70f private sector participation, 200­203 influences on road quality, 67­71, regulatory arrangements, 203­6 72­73f, 77 532 Index roads (continued) Rwanda poor, but improving, condition, civil conflict, 2­3 29­33, 30­31f, 32t, 34f exports, value of, 6 rural road conditions, 33­37, 35f, fuel taxes, 40 72­73f, 77­78 reform of institutions, 360 truck fleet conditions, 74­75 roads and corridors urban road conditions, 227­30, density, 20 228t, 229f over- and underengineering, 63 vehicle overloading and, 69­70, 365 rural roads, 22, 33 recommendations for future policies, spending, 55 77­79, 386­88, 388t trade clearance, 198 reforms at institutional level, 37­47, 79 road agency performance, 43­45 S rural road administration, 45­47 second-generation road funds, 29, SAA. See South African Airways 38­43, 39f, 42­43f, 78 SADC. See Southern African Development safety, 37, 79, 387­88 Community spending, 47­71, 48f, 77 safety and security budget execution, 52, 53f, 53t absence of safety culture, 362­63 capital investment, 50­56, 51f, 56f agenda for action, 385­86 maintenance expenditures, 56­60, airports and air transport, 151­52, 57­59f, 60t, 61f, 62t, 387 174­75, 176, 178n8 recommendations for funding as argument for national control of arrangements, 78 transport, 384 rehabilitation requirements, 54­55, ports and shipping, 210­12 54f, 55t roads, 37, 79, 387­88 road asset value as percentage of urban transport, 264­65, 267 GDP, 67, 68f Safety of Life at Sea Convention (SOLAS), road work costs and inflation, 60­67, 211, 223n9 65­66f São Tomé and Príncipe's oil exports, 194 sustainable financial basis for urban SARCC (South African Rail Commuter roads, 266 Corporation), 97, 231 urban transport, 227­30, 228t, 266 satellite technology, 142, 145 Road Safety Education Program seaports. See ports and shipping (Uganda), 37 second-generation road funds, 29, 38­43, Roberts, P., 279 39f, 42­43f, 78 RONET. See Road Network security. See safety and security Evaluation Tool Senegal Royal Air Maroc, 156, 158 bus rapid transit, 259 runways. See airports expenditures on transport infrastructure, and air transport 312, 318 Rural Accessibility Index (RAI), 33, 35f, railways, 4, 8, 87, 342 36, 279­80, 289 roads and corridors, 19, 29, 55, 58 rural standards of connectivity, 279­80, Sequeira, S., 358 289, 290t Setrag (railway operator), 115 rural transport Seychelles air traffic, 166 rail, 22 Sheltam (railway operator), 115 roads shipping. See ports and shipping administration, 45­47 Shyam, K. C., 279 conditions, 33­37, 35f, 72­73f, 77­78 Sierra Leone, collapse of, 360 connectivity model, 289 Sitarail, 9, 95, 113, 115, 122, 125, 131, critical to agriculture, 22 135n12, 135n14, 137n28 Index 533 Sizarail (Democratic Republic of Congo), Passenger Rail Authority, 366 107, 136n16 traffic trends, 92, 97 socialism, 3, 359 traffic volume, 87 Société Camerounaise de Transports residual funding gap, 347 Urbains (SOCATUR), 259 roads and corridors Société de Gestion du Patrimoine bilateral agreements on Ferroviaire du Burkina (Burkina operations, 76 Faso), 113 budget allocation, 52 Société des Transports Urbains du Congo concessions, 50b (STUC), 256, 263, 268n8 conditions, 29 Société de Transport Abidjanais (SOTRA), construction industry 259, 262 program, 375 Société de Transport en Commun over- and underengineering, 63 (SOTRACO), 251, 259, 268n7 paved roads, 20 Société d'Exploitation Ferroviaire des road density, 20 Industries Chimiques du Sénégal rural roads, 22 (SEFICS), 109, 135n14 safety, 37, 265, 388 Société Ivoirienne de Patrimoine spending on, 49 Ferroviaire, 113 toll roads, 340 Société Transgabonnaise (SETRAG), 101 traffic volumes, 26 Somalia, corruption in, 357 transport quality, 72 South Africa stock exchange raising capital, 344 air transport and airports trade routes, 19 domestic growth, 158 urban population, 308n4 international growth within Africa, South African Airlink Express, 145 160 South African Airways (SAA), 153, 155, ownership and management, 147 159, 160, 161, 172, 178n9 safety standards, 174 South African Civil Aviation bus rapid transit systems, 238­39 Authority, 151 corporate bond market, 343 South African Rail Commuter Corporation democracy and change, 14 (SARCC), 97, 231 expansion of public transport in South African Railways Association, 370 preparation for 2010 World Cup, southern Africa 239, 294, 311 ports and shipping expenditures on transport container traffic, 188 infrastructure, 315 customs reform and modernization, independence of, 2 198 local sources of finance, 343 dwell times, 216 logistics systems, 385 general cargo-handling performance, oil and gas exports, 194 217 ports and shipping railways capacity and demand, 208 competition with road sector, 105 container traffic, 182­83 concessions, 107 independent port regulator, 203 gauge lines, 86, 134n2 planning and reform, 195 roads and corridors private investing in, 340, 342 trade facilitation, 76 quality of governance, 357 transport price, 75 railways, 83­84 transport quality, 74, 74t, 76 commuter rails, 231, 238, 340 trade routes, 19 freight traffic, 98, 389 Southern African Construction Industry labor productivity, 117, 136n18 Initiative, 374 network structure, 86 Southern African Customs Union, 197 534 Index Southern African Development fuel taxes, 40 Community (SADC), 141, 151, local government networks, size of, 46 172, 179n12, 197, 198 ports and shipping Policy on Geometric Design of Rural customs, 197 Highways, 66 planning and reform, 195 Southern African Regional Construction private investing in, 342 Industry Council, 374 railways Southern African Transnet Freight Rail, 8 concessions, 115, 131 Soviet Union, spending on roads, 49 costs, 127 Spoornet, 86, 117, 134n4, 231. See now financial performance, 125 Transnet Freight Rail gauge lines, 86 strategic planning, 221, 236, 266, 395­96 passenger traffic, 95, 127 Sub-Saharan Africa regulation, 113­14 air transport and airports residual funding gap, 348 airport charges, 146 roads and corridors dominant airlines, 159 conditions, 29 traffic connectivity, 140, 140m, economic assistance, 397 154­55, 155f, 158 investment ratio to GDP, 55 trade performance, 6­10, 7m maintenance expenditures, 57 Sub-Saharan Africa Transport Policy over- and underengineering, 63 Program (SSATP), 18, 32, 38, 42, road funds, 42 44, 79n1 trade clearance, 197 Sudan Tanzania Railways Corporation (TRC), 87, air transport and airports, 160 92, 95, 112 oil exports, 193 Tanzania-Zambia Railway, 89 ports and shipping tariffs, rail freight, 101­4, 102­3f capacity and demand, 206 Tata Motors Ltd., 263 private investment, 342 taxes. See also fuel taxes rolling-stock productivity, 119 large bus sector, 250 rural connectivity, 279 minibus sector, 250 rural roads, 33 Tazara rail line, 84, 92, 119, 126 stock exchange raising capital, 344 Teravaninthorn, S., 74 Surface and Marine Transport Regulatory Togo Authority (SUMATRA, Tanzania), air traffic, 166 114, 262 roads and corridors, 19, 29, 33 sustainable arrangements for urban toll roads, 50b, 340 transport, 266, 370 Touré, Sekou, 360 Swaziland trade railways, 119 bilateral trade agreements, 75­76 roads and corridors, 29 ports and shipping, changing trade Swissair, 153 patterns, 181­95 syndicates. See unions and syndicates reduced due to transport conditions, 6­10, 7m as share of GDP, 11, 15n2 T traffic rates. See specific type of transport Tanzania Trans-African Concessions, 69, 343 airports Trans-African Highway, 19­20, 287 air traffic control, 141 Trans-Gabonais (railway), 84, 114 domestic traffic, 158 Transnet Freight Rail (South Africa) ownership and management, 147, 150 capacity and demand, 208 runways, condition of, 145 cross-subsidies, 384 economic development, 3, 92 labor productivity, 117, 136n18 Index 535 Metrorail and, 231 United States, energy interests in Africa, responsibility of, 366 193­94 traffic share, 87 urbanization transshipment, 190 history of, 6 Transparency International, 358b size and characteristics of cities, 225­27, Transrail, 92, 109, 113, 114, 115, 231 226t transshipment, 188­90 urban transport, 225­69 TRC. See Tanzania Railways Corporation bicycles, use of, 230, 238 truck transport. See roads buses. See bus rapid transit systems TunisAir, 156 institutions, 232­36, 233­34t, 266, 367 turboprop planes, 166, 174 impediments to integrated policy, 232­35 lack of strategy, 236 U motorcycles, 238 UATP (African Union of Public Transport), rails, 97, 231­32, 238, 268n3 370 recommendations for future policies, Uganda 265­67, 393­94, 395t air traffic control, 141 regulation, 260­65, 261t customs, 197 inadequate enforcement, 263­64 driver regulation, 264 principles vs. practice, 260­63 expenditures on transport infrastructure, reestablishing regulatory 315 framework, 267 local sources of finance, 343 safety, 264­65, 267 private investing in, 342 roads, 227­30, 228t, 229f, 266 railways congestion, need to reduce, 266, concessions, 112, 131 309n6 financing, 112 connectivity, 289­90, 289­90t track conditions, 88 land use and development, 235 roads and corridors safety improvements, 267 density, 20 sustainable financial basis for, 266 maintenance expenditures, 58 services, 236­50, 237t over- and underengineering, 63 availability and quality, 246­47, 246t safety, 37, 264 buses large and small, 239­41 trade clearance, 197 bus fleets, 241­45, 242­45t urban population, 308n4 costs, 249­50, 249t Uganda Taxi Operators and Drivers vehicle maintenance, 248­50 Association, 263 standards of connectivity, 280­81, UK Competition Commission, 161 289­90, 290t unions and syndicates, 247, 262­63, 365 strategic planning, need for, 266, 395­96 United Kingdom walking, 238 airport ownership, 150 U.S. Department of Transportation, 174­75 colonial interests of Great Britain in U.S. Federal Aviation Administration Africa, 2 Safe Skies for Africa Program, 151, 175, Transport Research Laboratory's 178n8 Overseas Road Notes, 66 standards, 150 United Nations U.S. National Transportation Safety Board, Conference on Trade and Development, 174 6, 212, 367 user charges, collection of, 323 Economic Commission for Africa, 6, 169, 268n10, 368, 370 V Food and Agriculture Organization, 36 Liner Code, 194, 223n5, 367 Victoria, Lake, 186b 536 Index W railways database, 277 ROCKS (road cost database), 62 WAEMU. See West African Economic and Urban Transport Indicators, 246 Monetary Union Worldwide Governance Indicators walking in cities, 238, 264­65 (WGI) project, 356 "Washington consensus," 355 World Economic Outlook (IMF 2010), 347 West Africa Worldwide Governance Indicators (WGI) air safety record, 174 project, 356 air traffic, 153, 166 bonded warehouses, 9 bus rapid transit, 239, 241, 244, 251 Y corruption, 357 Yamoussoukro Decision (1999), 167, ports and shipping 168­74, 179n11 capacity and demand, 206 cargo-handling costs, 218 Z constraints on economies of scale, 194 container traffic, 182, 188 Zambia customs reform and modernization, customs, 197 198 expenditures on transport dwell times, 216 infrastructure, 318 general cargo-handling performance, fuel taxes, 41 217 local government networks, size of, 46 lake transport, 186b local sources of finance, 343 transshipment, 190 national construction council, 374 railways railways competition, 106 concession financing, 130­31 concessions, 107 costs, 127 gauge lines, 86 financing, 112 regulation of commercial vehicles, 260 freight traffic, 98 roads and corridors locomotive productivity, 124 trade facilitation, 76 mineral extraction and, 89 trade routes, 19 private investment, 342 transport price, 75 regulation, 113 transport quality, 73, 74t, 76 roads and corridors West African Economic and Monetary conditions, 29 Union (WAEMU), 172, 179n12, maintenance expenditures, 57 369 over- and underengineering, 63 World Bank privately constructed roads, 45 air safety programs, 174, 175 rural roads, 33 bus rapid transit, support for, 235b safety, 264 concessional funding, 404 trade routes, 19, 184 development policy loans, 397, 405n3 Zambia rail company, 9, 106, 107 educational focus, 373 Zimbabwe funding tied to reforms, 359 air traffic control, 141 Global Economic Prospects, 14 private investing in, 342 Logistics Performance Index, 76 railways, 136n19 Private Participation in Infrastructure roads, spending on, 60 (PPI), 50b, 315­16 trade routes, 19 ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. The · 8 trees Office of the Publisher has chosen to print · 2 million BTU's of Africa's Transport Infrastructure on recy- total energy cled paper with 50 percent postconsumer fiber · 791 lbs of CO2 in accordance with the recommended stan- equivalent of dards for paper usage set by the Green Press greenhouse gases Initiative, a nonprofit program supporting · 3,809 gallons of publishers in using fiber that is not sourced waste water from endangered forests. For more informa- · 231 pounds of solid tion, visit www.greenpressinitiative.org. waste Transport infrastructure is a key requirement for economic growth. But in Saharan Africa today, road networks are sparse, conditions are poor, and accident rates are high. Railroads are in decline, and air traffic control and safety records are poor. Port performance falls well below international standards, and costs are high. Urban transit is generally chaotic and inadequate. The multistakeholder Africa Infrastructure Country Diagnostic (AICD) project provides unprecedented data on the nature, extent, and condition of transport infrastructure in the region. Africa's Transport Infrastructure: Mainstreaming Maintenance and Management presents, analyzes, and critically discusses these new findings against the institutional and policy background of the sector to identify the major causes of poor infrastructure performance. The analysis reveals that networks are generally less dense, both per square kilometer and per capita, than in other regions. However, because incomes are low, the financial burden is greater than elsewhere. As a consequence, maintenance has been neglected and the condition of the infrastructure has deteriorated, increasing the costs of operations. This situation is accentuated by failures in the management and regulation of operators. The book presents financing trends and compares them with estimated expenditure needs. It shows that backlogs in routine and periodic maintenance have built up needs for rehabili- tation that cannot be satisfied from domestic sources alone. It shows, both functionally and geographically, where major shortfalls are likely to occur and where priorities are highest, stressing the need to mobilize a wide range of sources for investment finance. The book concludes, however, that investment is only part of what is needed to improve the transport system. Unless that investment is accompanied by increased attention to maintenance finance and by substantial improvements in system management, the cost of transport in the region will continue to be a serious impediment to economic growth. This book will be of interest to governments in the region and to multilateral and bilateral aid and lending agencies, as well as to graduate students, faculty, and researchers in African studies and transport studies. ISBN 978-0-8213-8456-5 SKU 18456